Skip to main content

Full text of "Power railway signalling, by H. Raynar Wilson"

See other formats



No. 5. 

Railway Signalling 












Tel. Address 


Telephone NO. 3179, 






























Etc., Etc. 


| HEN the Author's Railway Signalling was published, power signalling and interlocking \vas in 
its infancy, and it was then his intention to write a second work upon the application of 
electricity to railway signalling. The use of power signalling grew rapidly, and a revised edition 
of Railway Signalling being soon called for, it was published under the title of Mechanical Railwtiv 
Signalling to which it was strictly confined and the Author decided to write a companion work to be 
called Power Railway Signalling, which should also include such matters as " lock-and-block," single 
line working, &c. 

This volume deals with what may be called the structure which has been raised on the foundations 
laid by the pioneers of the art of Power Signalling for railways, and among whom should be mentioned 
Cooke and Whetstone who introduced the speaking telegraph ; Edward Tyer, who enabled single lines 
to be controlled by electricity, thus greatly increasing their carrying capacity with safety ; \V. R. Sykes, 
whose " lock-and-block " is used all over the world ; Thos. A. Hall, who first convinced the railway 
world of the advantages of automatic signalling, and George Westinghouse who added to the benefits he had 
already rendered railways by demonstrating successfully the possibility of working points by other 
means than manual labour. 

The field covered by this work is, of course, a limited one, but it is one that is prolific of 
interest and highly specialised. Its importance is vast, for upon the accurate working of signalling 
mechanisms, probably more human lives in the civilized world are always depending than upon any 
other class of machinery, and the Author is satisfied that he has described every known system for 
working single lines, of operating automatic signals, and for the actuation of points and signals by 

In the course of five visits to the United States, the Author has been enabled to obtain most 
of his information as to American signalling first-hand and on the spot, and he has also personally 
inspected most, if not all, of the power installations on the Continent. 

The Author is greatly indebted to many friends, especially to his former colleagues in the 
railway world, also to the partners and managers of the several firms whose names are mentioned 
in the pages of the book, for the drawings, photographs and blocks they have placed at his disposal, 
and who have spared no trouble to render him all the assistance they could. 

Finally, the Author's thanks are due to Mr. S. R. Blundstone, Editor of The. Railway Engineer 
in the columns of which publication some portions of this book originally appeared for seeing the 
pages through the press. 




Manufacturers of, and Contractors for, 






Economical Working of Ordinary Semaphores 

with Primary Batteries, 


Lock and Block Apparatus 

Siemens Magneto System. 
Ferreira's Battery System. 



Siemens Mercury Rail Contacts or Treadles. 


Ra,iway Signalling Department { BROADWAY, WESTMINSTER, S.W. 




CHAPTER I. INSTRUMENTS. Single Needle Tver One- 
wire Tver Three-wire L. & Xorth-Western block 
Preece block Tver Three-position one-wire Povvles & 
Moore One-Wire Winter's block Pryce-Ferreira in- 
strument American block working- Mileage protected 
in America American block rules Switching-out boxes 
Telephone Light indicator Repeater Repeater for 
controlled signals Combined instruments - - Train 
describer Callaphone Apparatus for road level cross- 
ings Warning bells for level crossings Bell for warn- 
ing shunters Trains standing at signals Signalling 
long tunnels Key for intermediate Sidings Detached 
Vehicle Indicator Phonopore. pp. i 19 

Names of systems Early history Progress Objects 
Treadles and contacts Releasing key " Track-cir- 
cuits" as an addition or alternative Progress of "Track- 
cuits" Protection of converging junctions Con- 
trolling Releasing Key. pp. 20 26 

Systems of : Sykes Spagnolctti Hodgson Langdon 
Evans Blakey & O'Donnell Sykes, Jr., & O'Donncll 
Ferreira & Pryce Wood Siemens Bros. McKenzie 
& Holland Tver. pp. 27 47 

Progress in America Sykes' American Systems of: 
Coleman Patenall Winter McKenzie & Holland 
Siemens-Halske Sarroste & L-jppe Cardani. 

PP- 4857 

Hodgson Siemens Stevens & Sons Hollins Mercier 
Tyer McKenzie & Holland Cardani Position of 
contact maker Use of replacers Replacers of : Sykes 
- Tyer O'Donnell Hall Cardani Mechanical re- 
placers L'se of electrical detectors Detectors of : 
Sykes McKenzie & Holland Selectors Fouling, Clear- 
ance or Train Protection Bars Electric control of sig- 
nals Bonding rails Rail drills Insulated joints of : 
Weber Sykes Atlas Kohn. pp. 58 70 

Neath River Barnstaple Drumsna Severn Brey- 
don Ashton Barrow Suir American systems 
Chicago Boston Harlem. pp. 71 77 

SIDERATIONS. Early history Introduction of Tablet and 
Electric Train Staff Advantages Crossing orders in 
Great Britain Difficulties as to improved methods in 

America Progress made in America Outlying passing 
loops in America Controlled manual. pp. 78 81 

SYSTEMS. Tver's No. 5 Tver's No. 6 Tver's Auto- 
matic Tver's Automatic, with visual indications 
Tver's No. 5 permissive Tyer's Absolute Automatic 
McKenzie & Holland's Tablet Switching out Tablet 
Stations Banking engines Tablet pouch Exchanging 
tablets and train staffs at speed on : Highland R. G. 
Western R. Somerset and Dorset R. Unlocking start- 
ing signal by tablet or train staff Outlying sidings. 

pp. 82 96 

STAFFS. Webb & Thompson instrument Permissive 
staff Miniature staff Non-crossing stations Hansel's 
staff Unlocking starting signal Union Switch & Signal 
Co.'s staff. pp. 97 108 

AND OTHER METHODS. Progress in America Sugges- 
tion for British and Colonial lines Systems of : Winter 
Theobald Neale Hepper Sykes Illinois Central 
Dispensing with flagmen. pp. 109 122 

Early history Types Batteries "Track-circuits "- 
" Overlaps " - " Wireless-circuits " - - Three-position 
signals "Track-circuits" on electrically-operated roads 
Success on Interborough of New York Cost of Instal- 
lations " Normal-danger " and "Normal-clear" 
"Overlaps" Automatic train control Reasons for slow 
progress in Great Britain Installations in Great Britain 
and on the Continent Progress in America. 

pp. 123131 

stallation Hail disc Hall electro-gas Hall electro- 
motor Union Switch & Signal Co.'s disc Clockwork 
Electro-pneumatic U.S. & S. Co.'s Electro-motor Low 
pressure pneumatic Bezer's revolving General Electric 
signal Miller's signal " Three-position " American 
Railway Signal Co.'s signal Upwardly inclined arm 
Pennsylvania new system Staggering lamps. 

pp. 132151 

STEAM-WORKED RAILWAYS. Grateley-Andover Wood- 
head Tunnel Ashby Magna Alne-Thirsk In America 
Laroche-Auxerre St. Germain Bordeaux-Langon. 

pp. 152160 


Panton's illuminated arm Timmis-Laverazzi signal 
Waterloo and City Boston Elevated Brown's relay 
North Shore of San Francisco Struble's relay New 
York Interborough Metropolitan^ District Economies 
on District R. Illuminated diagram Magazine Train 
Describer Train Destination Indicator Underground 
Electric Railway Co.'s Tubes Great Northern & City 
Liverpool-Southport Metropolitan of Paris. 

pp. 161 184 

installation Harison signal Brecknell, Munro & 
Rogers' system Tierney & Malonc's automatic point 
mechanism Systems of : Electric Tramway Equipment 
Co. Siemens Bros. Siemens-Halske United States 
Electric signal Eureka signal Blake signal Philadel- 
phia and Western. pp. 185 192 

Switch locks "Track-circuits" for siding connections 
and crossovers Insulating point rodding, &c. Switch 
indicators Commutators Switch instruments Switch 
controllers Relays Signals for level crossings. 

PP- J 93 '99 

MATIC TRAIN CONTROLS. Purely mechanical not 
noticed Systems of : Boult Miller Sheehy Kins- 
man Laffas Jefcoate Raymond Phillips Western 
Syndicate Bonnevaile and Smith Raven On Electric 
roads. pp. 200 207 

PLANTS, THEIR PURPOSE. Early history Disadvantages 
Advantages Doubt as to economy of time in operation 
and of signalmen " Track-circuits " in lieu of or in addi- 
tion to locking bars Power for working gates, also for 
signals at mechanically-operated boxes Power-worked 
distants Itinerary levers Summary of power plants in 
use. pp. 208 213 

Boston Southern Dwarf signal Signal movement 

Point movement American locking frame Installations 
at : Long Island St. Louis Pittsburg Thompson 
Bishopsgate Bolton British locking frame Electro 
pneumatic slot British point movement Power- 
operated gates Installations at : Hull Tyne Dock 
Metropolitan District Underground Electric Railway 
Co. 's Tubes Newcastle-on-Tyne Glasgow Cairo 
Howrah Siemens-Halske electro-pneumatic gates. 

pp. 2!4 238 


Grateley Salisbury Staines Woking-Basingstoke 
- Ardwick - Newton -- Clapham Junction -- Locking 
frame Point movement Signal movement Automatic 
restoration of signals Selection of signals Signals 
operated automatically when box closed Wath Push 
button machine Route indicator Ermont Bogue & 
Mills pneumatic gates. pp. 239 254 

" Crewe " Crewe L. & N.W.R. Severus Junction 
Siemens-Halske Antwerp Siemens Bros. Derby 
Didcot Siemens Bros.' new system Snow Hill, Bir- 
mingham Taylor Cost of maintenance Sixtieth and 
Clark Streets, Chicago South Englewood New York 
Central Oakdale British Taylor Point movement 
Signal movement Locking frame Union Switch & 
Signal Co.'s locking frame and switch movement 
Bleynie, Ducousso & Rodary system Sykes' switch 
movement Johnson's system. pp. 255309 

St. Enoch's, Glasgow Victoria Station, Pimlico Elec- 
tric illuminated signal Working of points Locking 
bars not attached to plungers Locking frame Sema- 
phore motor Detectors Plunger detectors. 

Servettaz Progress in France Installation at Cape 
Town -Nagari Locking frame L'Aster Dcscubes' 
system. pp. 3 10 333 

APPENDIX. Powles and Moore's Three position one-wire 
block instrument Pangbourne and Goring signals 
E.P. plant at Glasgow Central E.P. constant detection 
Rules for automatic signals. pp. 334 339 



A companion work to this. The only complete text 

books on the subject. 



Being the 2nd Edition, much enlarged and revised, of H. Raynar Wilsons " Railway Signalling'' Royal 4/0 
(12! ins. by IO ins.), bound in cloth. 200 pages. Nearly 500 illustrations, mostly scale drawings, 

and 3 folding plates. 


great care has evidently been bestowed on its production .... the author has thorough knowledge of his subject." Tkt Engineer. 

. necessarily of technical value to the profession." The Railway Nnvs. 

. is altogether the best that has ever appeared on the subject of signalling." American Engineer and Railway Journal. 
Each subject treated in great detail." The Railroad Gazette. 

covers the field of mechanical signalling." The Railway and Engineering Rc-.'i, ;<. 

offers to those interested in mechanical railway signalling a thorough compendium of British practice which those interested in the operation of railways, 
no matter of what country, will find of great interest .... Some useful suggestions for signal specifications and a very convenient index concludes 
this valuable contribution to railway signal literature. The paper is of good quality, the illustrations are of a high order and the typography is 
unexceptionable.'* The Rail-may Ant. 

Chapter I. 
Chapter II. 
Chapter III. 
Chapter IV. 
Chapter V. 
Chapter VI. 

Chapter VII. 

Chapter VIII. 
Chapter IX. 

Chapter X. 
C/tapter AY. 


INTRODUCTION. Progress "Readable" signals Splitting^ distants Mishaps at facing points 
Increase in interlocking Signalling in America. (34 illustrations.) 

SINGLE LINES. History Wooden train staff Intermediate sidings Locks for intermediate 
sidings. (16 illustrations.) 

SIGNAL BOXES. Wood or Brick Details of construction -Types of different companies 
Overhead cabin Furniture and fittings. (44 illustrations.) 

SIGNALS. Wood and Metal Preservation of post Arms and fittings -Lamps Slots Signal 
bridges Route indicating signal Dwarf signals. (87 illustrations.) 

POINT AND SIGNAL CONNECTIONS. Wire Wheels Cranks Rodding Joints Facing Point 
locks Locking bars Protection bars Rollers Detectors. (150 illustrations.) 

LOCKING FRAMES. Railway Signal Co. Lancashire & Yorkshire Stevens Saxby & Farmer 
Dutton London & North Western McKenzie & Holland Tver Evans, O'Donnell Great 
Western Selectors. (39 illustrations.) 

SIGNALLING PLANS AND LOCKING LISTS. Position of cabin, signals and points Maximum 
distance for points Distant, home, starting and siding signals Crossover junctions Short 
sections Sighting signals Single line passing places Interlocking table of 7 typical places. 
("24 illustrations.) 

LEVEL CROSSING GATES. Details for simultaneous working Gate control Wickets. 
(25 illustrations.) 

LARGE SIGNALLING INSTALLATIONS. Diagrams of signalling of Euston ; Edinburgh (Waverley); 
Liverpool Street ; Newcastle-on-Tyne ; Holloway, with locking list ; Belfast ; Dundalk ; Manchester 
(Victoria), with locking list ; Blackpool (Talbot Road), with locking list ; Glasgow (St. Enoch) ; 
Reading West Junction ; Waterloo. 

SIGNAL DEPARTMENT. Administration; Procedure; Staff; New Works ; Maintenance; Rules. 

Price, i8j- Post free in United King Join, i8j6. Abroad, 19 /- 
Order direct from the Publishers of ' The Railway Engineer," 

3, Ludgate Circus Buildings, 


(See also page xiii. for other publications.). 






o - 

2: o 


J X 
O S3 






- c 

^ W 


U w 

2 o 


H " 









THERE is not much variety in the form of block instru- 
ments used by railway companies. Some of these are illus- 
trated in Chapters II., III. and IV., dealing with " Lock- 

A very popular system is to provide a bell instrument and 
a separate block instrument for each line, or three instru- 
ments for each direction, i.e., an instrument for trains to be 
sent, another for trains approaching, and a bell instrument 
for both purposes. 


Another form is a 
combination instrument 
where the signals for 
one line are given on the 
upper part of the instru- 
ment, for the other line 
on the middle of the in- 
strument, and the ringing 
key is at the bottom. 

There are also " One- 
Wire " and " Three- 
Wire " instruments. In 
the former, the bell sig- 
nals and the indications 
for both up and down 
trains pass over the 
same wire. In the latter, 
a separate wire is pro- 
vided for the up-signals, 
for the down-signals and 
for the bell. It will be 
apparent that the "Three- 
\Yire " system is the 
better, but it is far more 

costly to instal, and advocates of the " One-Wire " system 
can claim that the " One-Wire " instrument is quite as safe 
as the " Three-Wire." 

Single Needle Instrument. 

Fig. i represents the popular single needle instrument 
which is used where separate instruments are provided for 
up and down trains. The needle is normally in the centre as 
shown and indicates " line-blocked," i.e., closed. When a 
train has to be sent say from A to the next box B, the signal- 
man at A gives so many beats on the separate bell instru- 
ment fixed near the block instrument, when, if the signal- 


Fig. i. Single Needle Pegging 
Block Instrument. 

man at B can accept the train, he will turn the commutator 
handle to the left, so causing the needle of his own instru- 
ment and the corresponding instrument at A to point to 
" line-clear." It is kept in this position by a peg (see illus- 
tration) which is put through the spindle of the handle. 
When the train leaves A the peg is taken out and the com- 
mutator turned to " train-on-line " when it is again pegged in 
the new position. 

Tycr's " One-wire " Instrument. 

Fig. 2 illustrates 
Tver's " One-Wire " in- 
strument, with miniature 
arms for indicating the 
position of the section 
ahead. The instrument 
is bell and block com- 
bined, the bell being in 
the upper portion. The 
bell plunger and commu- 
tator disc are placed 
together, the spindle 
for the bell being- inside 
the shaft of the commu- 
tator. The indications 
given on the screen above 
the commutator are 
" line-clear," " train-on- 
line " and " train-out-of- 
section," the last named 
being the normal condi- 
tion. When B accepts a 
train from A he turns his 
commutator so as to 
give the signal "line- 
clear," and this lowers the lower arm of the miniature sig- 
nal on his instrument and the upper arm of the instrument 
at A. When the train leaves A the prescribed bell signal is 
given by A to B and the man at B then turns his commu- 
tator to " train-on-line " (in which position the commutator 
becomes mechanically locked), and on pressing the bell 
plunger to acknowledge the signal both arms are raised 
again. When the train arrives at B the signalman there 
gives the prescribed bell signal and presses in the small 
plunger on the right-hand side, which causes " train-out-of- 
section " indication to appear. 

Fig. 2. Tver's One Wire 
Block Instrument. 

P O W E R R A 1 L W A Y SIGN ALLl N G . 







''"' 3- Iyer's Three Wire 
Block Instrument. 

Tycr's " Tlircc-wirc " 


This is illustrated by fig. 3. 
The upper dial A is for trains 
that are being sent to, and the 
lower dial is for trains that 
are coming from, the next 
box. The bell plunger is 
marked D and the commu- 
tator disc B, and they arc 
similar to those in fig. 2. The 
indications C are those given 
to the next box and show 
when the man has given 
" line'dear," " train-on-line " 
and " line-closed. " 

L. and 'N. W . Instrument. 

In fig. 3 is illustrated the block instrument used on the 
L. and North Western R. As will be seen, each instrument 
indicates the block signals sent and received whilst the bell is 
in the lower part. 

Three views are shown. The one on the right represents 
the instrument as fixed on the shelf ; in the centre is a side 
elevation and on the left is an internal view with the door 
open and swung back. 

When the bell a is rung current enters from line from con- 

tact b to the coils c. The magnet is thereby energised and 
the armature attracted so that the hammer / strikes the bell. 
When a bell signal has to be sent to the adjacent box the 
ringing key g is depressed, which breaks contact at h and 
makes at h?, joining up line to the local battery, whereby bell 
signals are sent to the other box. 

After the passage of the necessary bell signals, should the 
signalman wish to accept a train he would do so by the com- 
mutator j. In the left-hand illustration the back of the com- 
mutator is seen. It has two plates k fe 2 , and there are four 
springs /, P, / 3 , I*. The former two, I P, join up the battery, 
spring / 3 joins up line and spring I* goes to earth. The com- 
mutator is kept in position by the latch m entering between 
the pins n, it 1 , n 3 , n 4 , the latch being held down by the spring 
a. The correct movement of the commutator is obtained by 
the pins n n* coming against the stop p. When the " line- 
char " signal has to be sent the commutator is turned to the 
left (to the right as seen in the left-hand illustration), conse- 
quently the plate fe joins up springs / 7 4 and P P, which causes 
the needle in the box in advance and the needle of its own 
instrument to be deflected to "line-clear." When the "traih- 
cnterinrr-section " signal is received the commutator is 
turned completely from left to right, so that / P are joined 
and / 3 / 4 , which reverses the needles, so that they are deflected 
to " train-on-line." \Vhen the train has passed the commu- 
tator is turned to normal, so that the battery is disconnected, 
and the needles, being balanced, go to the central position. 

The screw q is for regulating the stroke of the bell ham- 

Fig. 3A. Combined Block Instrument; London and North Western Railway. 


met; the spring r is for keeping the ringing key in position, 
and the shutter s is to give access to the axle of the ringing 

A flexible wire is run from the spring I to the strap t of 
the lightning arrester. 

Preece's System. 

The Preece instruments, whether on the " One-Wire " or 
' Three-Wire " method, consist of a miniature semaphore 
arm which, when up, indicates, like a fixed signal, danger, 
i.e., that the section ahead is not clear. When the arm is 
down it signifies that a train may enter a section. The com- 
parison with fixed signals goes still further. The commutator 
for working the signals is like the lever in a locking-frame, 
and is back when the section is closed, and over when it is 
open, so that when a signalman gives permission for a train 
to approach him, he pulls the electric switch lever over, and 
his permission to send a train is given by the lowering of the 
miniature semaphore. The bell instrument for sending the 
prescribed bell code is provided with an indicator showing 
what signals have been sent to the next box, and the elec- 
trical connections to the switch lever and the bell are inter- 
locked with each other so that they must agree. 

Tver's Three-Position One-ll'ire Instrument. 

Mr. G. P. \eele, in his reminiscences, states that when 
the question was submitted to the late Sir Richard Moon as 
to whether the L. and North Western R. was to be equipped 
with block instruments worked by one or three wires, the 
economical spirit that was so strong in the mind of that railway 
magnate made him determine that the " One-Wire " system 
was to be adopted. It is very clear to anyone that there must 
be a large saving in the running of block telegraph wires 
when only one wire is used instead of three, but hitherto there 
has been one drawback, which is that in " One-Wire " instru- 
ments of the Tver and Preece type, where miniature signal arms 
are used to indicate the state of the line, that there can be only 
two indications given by the arm, viz., "line-clear" when 
the arm is down, but the upward position of the arm has to 
represent "train-on-line," also the normal position of the 
instrument " line-closed." 

In the instrument illustrated by fig. 4 this difficulty has 
been overcome by Tver and Co., Ltd. It is worked by one 
wire, and yet gives by the position of the needle the three 
indications, viz., " line-blocked " (the normal position of the 
instrument), " line-clear " and " train-on-line." 

It will be seen that the instrument combines what is neces- 
sary for the up and down lines, i.e., for trains that are sent 
to the next box on one side and for the trains that are received 
from the same box. There is one common bell which is fixed 
at the side or above the block instrument. The lower indica- 
tions given by the screen are the block signals that the man 
has sent to the next box. 

The ringing bell a is fixed on a spindle that passes through 
the armature b as in fig. 2. On the lower left hand side is a 
small button c which, on being pressed in, frees the commu- 
tator and allows it to be turned. It cannot, however, be 
turned until the bell plunger has been pressed in. By this 
means it is ensured that no alteration is made in the position 

Fig. 4. Tyer's Une-u ire Three-l'osition Block Instrument. 

of the block needles without the bell ringing and attracting 
the attention of the man at the next box. 

The dial movements are of the Tver cylinder type and are 
mounted on the dial plate, and in close proximity to each is a 
small local locking coil, having at the extreme end of its arma- 
ture a small cam escapement. This cam is so arranged that 
it allows of a free traverse of the cylinder armature and needle 
in one direction, but it checks the reverse movement and holds 
the needle when half the distance has been completed. The 
lower of these local locking coils engages with the armature 
of the needle recording the outgoing currents and it is 
operated by the commutator and local battery. The upper 
locking coil engages with the armature of the needle record- 
ing received currents and is operated by a special relay in 
the line current and which is fitted in the bell case. This 
relay has two pendant armatures, one responding to cur- 
rents of full strength and the other to those of fractional 

In the normal position of the instrument, as illustrated, 
the needle shows " line-blocked " and the first commutation 
from this position sends a current of full strength to line 
which causes the needle to traverse the full distance and to 
give " line-clear," at the same time operating the heavier 
armature of the line relay which simply closes the bell circuit. 
The next commutation sends a reverse current of full strength 
to line which also operates the heavier armature of the line 
relay, but does not allow of a complete traverse of the 
needle in the reverse direction, owing to the cam of the small 
locking coil checking the cylinder and allowing the needle to 
occupy the central position " train-on-line." 

The third and final movement of the commutator splits the 
current and short circuits three-fourths of it through the lower 
locking coil. This further movement operates the armature 


allowing the needle recording outgoing currents to show 
"line-blocked." At the same time the remaining current of 
one-fourth the full strength is sent to line and this current 
being too weak to operate the heavier of the line relay arma- 
tures, immediately attracts the lighter. This, in addition to 
closing the bell circuit, operates the upper locking coil, the 
armature of which engages with the cylinder of the receiving 
needle and the latter is freed and allowed to traverse the 
remaining distance and to show " line-blocked " which is the 
normal position. 

An advantage of these instruments is that they can be 
readily converted for " Lock-and-Block " at a very small cost. 

They have been adopted as the standard instruments on 
the North Stafford R., they are in use on the Great Eastern 
and other railways in Great Britain and on the Cape Govern- 
ment Rs., whilst on the Furness R. the instruments now in 
use are to be replaced by those of the type described and 

Powles and Moore's One-Wire Instrument. 

The Walters Electrical Manufacturing Co., London, 
have a " One- Wire " block instrument, which the firm manu- 
factures in considerable quantities, chiefly for Foreign and 
Indian railways. 

It will be seen from the external view, fig. 5, that the 
appearance of the instrument is similar to that of ordinary 
block instruments. 

The upper or dial portion consists of two electric minia- 
ture semaphore arms. The upper one represents the starting- 
signal, and is worked from the box in advance. The lower 
arm represents the starting-signal of the box in the rear, 
for a train coming from that direction. The upper semaphore is 
manipulated by the signalman in the advance box, and cannot 
be altered in any way by the commutator of its own instru- 
ment. The lower semaphore is worked by the outgoing cur- 
rent sent by the commutator of the instrument, and is syn- 
chronous in its movement with the upper semaphore arm on 
the instrument in the rear box. 

The visual signals being: in the form of miniature sema- 
phores, the signal received in reply to " is-line-clear " is un- 
mistakable, as the semaphore arm either remains at danger 
with the answering bell signal, or is lowered to " line-clear " 

The movement of the miniature semaphore therefore leads 
the position that the outdoor starting-signal should be put to. 

A signal sent when the commutator indicates either " line- 
closed " or "train-on-line," puts its own lower miniature 
semaphore, and the rear cabin's upper miniature sema- 
phore, to the danger position, but when the commutator 
indicates " line-clear " and a signal is sent, these semaphores 
are lowered. The semaphores cannot be moved without 
giving an audible signal on the bell in the distant box. 

The state of the section is indicated clearly to the man in 
the advance box in clear bold type, coloured as follows : 

Line-closed black letters on white ground. Line-clear 
white letters on green ground. Train-on-line white letters 
on red ground. 

The commutator is arranged to show either of three indi- 
cations, the one visible denoting the exact state of the section 
which the instrument governs. 

Fig. 5. Powles and Moore's One-Wire Block Instrument 
l< Section A B 

A- Is Line Clear? 


8 - Moves Commutator 
& Kings 



A - Lowers Outdoor 

A- Gives Tram entering, Section i 


fl- Moves Commutator 
* Kings 

A - Raises 

Tramarrives at B 
& passes out of Section 
B - Moves Commutator 
& Rings LINE- 


A- Gives Arrival 'Acknowledgement 





- Black on White. 

- White on Green. 

- White on Red. 


It is compulsory, when the signalman has given " line- 
clear," that he must also send " train-on-line " before he can 
bring the instrument back to its normal position of " line- 

For receiving trains the commutator works in a complete- 
cycle, and once it has been turned to indicate " train-on- 
line," the instrument cannot be put to " line-clear " until 
" line-closed " has been sent. 

The instrument may be put to " train-on-line " from cither 
the " line-closed " or " line-clear " positions. 

Directly the commutator is moved from one indication to 

another, it is locked in that position, and cannot again be 

altered until the plunger has been depressed, thus setting the 

Fig. 7. Fig 

The construction and working of the commutator is as 
follows : 

The plunger b 3 moves in a spring box b l , which rotates in 
a bush b fixed on to the door of the instrument, fig. 10. The 
spring box b'' is rotated by means of the milled finger piece />- 
and which in turn rotates the plunger by means of the pin d-, 
which slides, when the plungei is plunged, in a slot d- cut in 
the spring box b 1 . 

Towards the end of the plunger is fixed a large insulating 
block c 1 on to which are fixed the contact arms c- c 3 , which 
make the necessary electrical connections between the pairs 
of contact springs, k k l fe 2 k 3 fc 4 fc 5 . Into the end of the 
plunger rod is screwed an insulating cap d, which carries a 

Fig. 1 1 

Fig. 8. 

electrical semaphores according to the indication shown by 
the commutator, and also ringing the bell at the other box. 

The bell plunger is always free to be used. 

The electrical commutator being circular requires the 
minimum number of contacts to be made, and is so arranged 
that it is impossible to short circuit the battery. 

The whole of the exterior metal work is always at earth 

The ordinary code of block working can be adhered to. 

Only one line-wire is required, and only one battery of 
from 4 to 6 Leclanche cells is necessary for each instrument. 
The sending line battery also acts as the local battery in the 
relay bell circuit. 

Fig. 6 shows diagrammatic-ally the working of a pair of 
these instruments. Figs. 7 to 13 show the internal mechanism 
by which the results enumerated above are obtained. It is 
strong and simple, and brass spring rubbing contacts are 
used throughout for the electrical connections no silver or 
platinum contacts beiny; used. 

Fig. 12 

contact piece d 1 which norm-Ily makes a connection between 
the contact springs fe 6 , but when the plunger is plunged the 
springs pass off d 1 on to d and the connection is broken. The 
contact arms c 2 r* are of different widths and the guard ring 
k ? , which is a little higher than the contact springs, has three 
pairs of slots cut in it so that the plunger can only be plunged 
when it is in one of the three correct positions. The guard 
ring is cut at fe 8 fe 9 , fig. 9, so that should the arms c 2 c 3 rub on 
it no electrical connection is made. 

The indicator plate d 4 , figs. 8, 10, and 1 1, is rigidly attached 
to the spring box b 1 and it is locked in position by the bolt 
g", which is carried on the contact arm r 3 and in a rectangular 
hqle g, with the indicator plate d*. The bolt carries a spring 
in compression which normally tends to shoot it through the 
indicator plate, but this is prevented by a shutter gate g l 
carried on the indicator plate by a bolt g-, on which it is free 
to oscillate, but is maintained in its central position, closing 
the hole g, by two spiral springs g*. The upper part g* of the 
shutter is cam shaped, and when the indicator plate is rotated 


Fig. 14. 

it comes into contact with one of three fixed pegs h 4 , /i 5 , h*, 
and is therefore pushed aside and allows the bolt to shoot 
through into one of three notches /z 1 , 7^, /J 3 , cut in the fixed 
ring h, so that the indicator plate cannot be rotated again 
unless the bolt be withdrawn by plunging, and thereby giving 
an indication to the other signal box. Directly the bolt is 
withdrawn the hole g is closed by the shutter g l . The rota- 
tion of the indicating plate in the direction to give the proper 
sequence of signals, viz., "line-clear," "train-on-line," 
" line-closed, " is secured by fixing a spring pawl e l , so that it 
always presses on the edge of the plate and having its detent 
so shaped that it will allow the pins e 7 , e*, on the back of the 
plate, fig. 8, to pass it in one direction. But in order that 
the indication may show " train-on-line " from either of the 
other two indications the third pin is omitted. 

The arrangements of the circuits and working can be readily 
understood by referring to fig. 12. A and B shows dia- 
grammatically two sets of complete apparatus, A being at 
one end and B at the other end of a section. Suppose that a 
train is required to be sent from A to B, and is standing at 
A, both the up and down lines being closed. Both indicators 
show " line-closed," and all miniature semaphores are at 
danger position. A asks B for permission to send a train 
by pressing his plunger. This rings the bell at B, the cir- 
cuits being as follows : Starting from copper pole of bat- 
tery m, the current goes through the contact maker which 
connects k pair of springs to " line," and along same to 
instrument at B- Here it goes across springs fe 6 to coils 
m 4 m 4 , thence though the relay coils w 5 w 5 to earth, 
back through earth to instrument A again, through coils 
m'm 1 , across contact maker between springs fe 3 , and thence 
to zinc pole of battery m, thus completing the line circuit. 
This current on going through the relay coils rn 5 m 5 of instru- 
ment B, attracts the armature m 5 , and closes the local cir- 
cuit which energises the coils m 7 m 7 , which rings the bell. 
As the indicator at B is showing "line-closed," he knows 
that there is no train in the section. He then rotates the 
indicator plate to show " line-clear," and then presses in his 

Fig. 13. Powlos and Moore's One-\Vire Block Instrument. 

plunger. In this position, as previously mentioned, when 
the plunger is pressed, springs /e 6 are broken, and fe 2 and fe 5 
respectively connected together. The electrical circuit is 
thus made similarly as explained with instrument A, but in 
this position it will be noticed that instead of the copper 
pole being put to line and zinc to earth, zinc is put to line 
and copper to earth. Thus the bottom flag p at B and the 
top flag p l at A will be lowered, and the bell at A rung. A, 
now seeing that his top miniature semaphore is lowered, 
knows that he may lower his outdoor starting signal, which 
he does, and the train proceeds towards B- Directly the 
train has entered the section, A informs B that this is so by 
means of a certain number of beats on bell at B- B then 
rotates his indicator plate to show " train-on-line " and 
plunges, that is copper is put to line and zinc to earth. This 
raises the miniature semaphores p and p l , and rings the bell 
at A- The train is now in the section proceeding towards B, 
and the indicator at B is showing " train-on-line." When 
the train arrives at B, and passes out of the section, B 
rotates his indicator plate to show " line-closed," and 
presses his plunger the requisite number of times to give 
the code signal on the bell at A, and as the battery current 
is in the same direction as for " train-on-line," the miniature 
semaphores remain at danger. 

These instruments have also been adapted for use on 
single lines and at junctions. 

When used on single lines, a make and break switching 
arrangement is added, so that when the instrument at A has 
its indication showing anything else but " line-closed " the 
top miniature semaphore coils are short circuited and no 






0^ 00 







current can energise them. The way in which this is done is 
easily followed on fig. 14. The normal position of spring q l 
is to press against contact screw in q-. On the end of the 
contact arm c 3 is a pin q of sufficient length to press down 
spring q 1 and break connection with q- when in the " line- 
closed " position, and on the contact arm c 3 being moved 
from the "line-closed" position, the spring rises and 
makes contact with q-, and thus (the electrical connections 
being added as shown) it follows that the coils m- are short 
circuited when the spring is in its normal position, and free 
to be energised when the spring is pressed by the pin on the 
end of contact arm c 3 . Therefore when A has once given 
permission for B to send a train, and moved the contact arm 
r 3 , the miniature semaphore /> l (which leads the outdoor 

semaphore for trains to B) cannot be lowered by B until the 
train from B is out of the section, and the instrument put to 
" line-closed." 

For use at junctions, the arrangement shown in figs. 15 
to 17 is provided. Fig. 16 shows the extra electrical connec- 
tions necessary for three sets of instruments in one signal 
box. The electro-magnets when energised attract an arma- 
ture s l and lift a catch s- working in s 3 between the two 
catches s l which are fixed to indicator plate d 4 , thus locking 
the indicator plate and preventing any rotary motion. 

A similar make and break switching arrangement to that 
above described, fig. 14, is provided, with the addition of a 
bottom contact q 5 . This is also worked in the same manner 
by means of a pin q in contact arm c 3 . Thus when " line- 
closed " is indicated, contact spring 5 s is making contact with 
q 5 , but at either of the other two indications it breaks with 
q 5 and makes with contact cock s 7 . Therefore when any one 
of two or more instruments thus connected up is moved to 
" line-clear " or " train-on-line," the electro-magnet circuits 
of the other instruments will be complete, the armatures and 

Powles and Moore's One-Wire Block Instrument. Fig. 17. 

catches of same will be raised, and the rotary motion will br 
locked and cannot be moved from " line-closed " until the 
other instrument is put to the position " line-closed." 

Where junction and single line working are required 
together, the working is combined in the one switching 
arrangement as shown in fig. 17. The miniature semaphore 
coils m- are short circuited through the contact springs it' it' 1 , 
which are insulated from the spring plate it- 5 . 

Moore and Powles' AVw One-Wire ^-Indication Instru- 
ment is described and illustrated in Appendix A, p. 334. 

Winter's Block Instrument. 

Fig. 1 8 shows Winter's instrument for double lines. 

It is the practice in India to hand the driver a "line-clear" 
ticket as his authority to proceed. These are kept in an 
instrument, and access is obtained to them by means of a 
small drawer i, which when pulled out is found to contain 
one ticket. 

It is necessary to control this drawer so that it cannot be 
pulled out without the permission of the man at the next 



'i::' !; .n 

.LINE . 

Fig. 18. Winter's Block Instrument. 


Fig. 19. Kleclrical Connections, Winter's Block Instrument. 

station. This is done by means of the electrical connections 
shown in fig. 19. The drawer has two racks, a e, which are 
engaged by the pawls F, G, the latter is to prevent the drawer 
being pulled out without the permission of the next man, and 
the other pawl F is to prevent the drawer being put back until 
the next man has sent the " train-arrived " signal. .There is 
the usual bell instrument P, a button K, and a switch lever 
H. This lever is held in the normal position by the lock M, 
and in the over position by lock M 1 . When a train has to be 
sent the plunger P is pressed, which rings a bell only ; but 
if at the same time as the plunger P is pressed the button K 
is also pressed for a prolonged beat, the lock M is attracted, 
and the lever H may be moved from "off" to "on." This 
reverses the line battery, and the relay tongue at the distant 
station goes from T to S, so that when the plunger P and 
button K are again pressed, the electric magnet of the pawl 
G is energised, so unlocking the ticket drawer and allowing 
it to be pulled out. 

The drawer remains out until the " train-has-arrived " 
signal is received from the distant station, which signal can- 
not be given until the train has arrived, and here there is a 
combination of " Lock-and-Block. "' At the distant station 
the signal wire is coupled to the slide a, fig. 20, which carries 
the cam b. When the signal wire has been pulled sufficiently 

far to lower the signal, the projection b 1 comes against the 
projection c 1 of another cam c, and raises the latter. This 
frees a third cam d, which falls under cam c, and at the same 
time raises trigger e to level with the top of the rail. The 
pulling of the slide a brings springs f f into contact, and 
these are in the releasing circuit, but the circuit is not com- 
plete until the pin g on the cam c comes into contact with 
springs h l h?, which happens when the trigger e is displaced 
by the arriving train. That completes the circuit, and the 
electro-magnet on the pawl F (fig. 19) is energized and allows 
the drawer to be closed, and the switch lever returned from 
" on " to " off." 

Pryce-Ferreira Block Instrument. 

One feature in this instrument is that the signalman can- 
not change the needles of the instrument from one position to 
another without ringing the bell and so attracting the atten- 
tion of the other man. 


Fig. 21. Pryce-Ferreira Block Instrument. 

Fig. 20. Releasing Treadle, Winter's Block Instrument. 

Fig. 22. Details of Pryce-Ferreira Block Instrument. 


Fig. 23. Commutator 
Connections, Pryce- 
Ferreira Instrument. 

Fig. 24. Pryce-Ferreira 
Interlocked Block Instrument. 

Fig. 25. Commutator 
of Pryce-Ferreira Inter- 
locked Block Instrument. 

A second feature is the greater sweep given to the needles, 
so as to avoid any ambiguity as to the position they are in. 

A third feature is the interlocking that may be applied 
between the two instruments, applicable to converging lines 
at a junction. 

The instrument is illustrated by fig. 21, and the internal 
details by fig. 22. The ringing bell is c, which when pushed 
in moves the contact lever e 1 , and moves the contact points 
<>' e 9 from spring e 1 to springs e* e*. so ringing the bell in the 
other signal-box. The commutator for moving the indicating 
needles is b, the handle outside the instrument for moving 
the commutator is b 1 , which is coupled to the spindle b-. The 
spindle has a slotted arm c 1 engaging with a pin c- on the 
moveable part b- 1 of the commutator. When the commutator 
is turned to the right or to the left, different circuits are set 
up (as will be understood from fig. 23) which cause the needle 
in the adjacent signal-box to show " train-on-line " or " line- 
clear " as the case may be. At the same time the indicator 
plate d attached to the hub c on the spindle shows at d 1 what 
signal has been sent. 

In order to prevent the commutator from being moved 
without ringing the bell the stem of the ringing key e has at 
the far end an arm e 3 , to which is fixed a pin e 4 , which passes 
through suitable openings in the hub c into a locking plate e 5 , 

Fig. 26. Electrical Connections, Pryce-Ferreira 
Interlocked Block Instrument. 

fixed to the inner side of the instrument. This holds the hub 
c fast, and consequently the commutator cannot be moved 
until the ringing key has been pressed in sufficiently far to 
carry the pin e* free of the locking-plate, and this cannot be 
done without moving the bell contact lever e 1 . The lever / 
with the notches f f 1 engaging with the pin f on the hub c, is 
to guarantee the holes e 6 t 6 on the locking-plate being in posi- 
tion to receive the pin e 4 when the ringing key is released 
from the signalman's pressure. 

At a junction where two lines converge say, from Y and 
Z, meeting at X the instrument applicable to the line from 
Y is interlocked with the instrument from Z and vice versa. 

This is done by providing in each instrument a separate 
electro-magnet which is energised by the current of the other 
instrument. This magnet is 5 in fig. 24, the armature of 
which is s 1 . When the armature is attracted, it holds down a 
lever t l engaging in a notch in a special disc t attached to the 
spindle of the commutator of the block instrument. 

The arrangement of the circuits in this case will be seen 
in fig. 25 and the working will be understood from following 
the wiring diagram in fig. 26. 

American Block Working. 

As understood in Great Britain there is very little block 
working in America. Most of what is called " block-work- 
ing " is effected by messages exchanged between the opera- 
tors (or between operators and the train despatcher) on tele- 
graph instruments.* 

The following is the block code adopted by the American 
Railway Association : 

31 1.\. The normal indication of home and ads'ance block signals is 
" Stop." . ,, 

3116. The normal indication of distant block signals is " Caution. 

312. Signals must be operated carefully and with a uniform move- 
ment. If a signal fails to work properly its operation must be dis- 
continued and the signal secured, so as to give the normal indication 
until repaired. 

313. Signalmen must observe, as far as practicable, whether the 
indication of the signals corresponds with the position of the levers. 

314. Signalmen must not make nor permit any unauthorised altera- 
tions or additions to the apparatus. 

315. A block record must be kept at each block station. 

316. The prescribed telegraph signals are as follows: 
i. Display stop signal. Answer by S D or 5. 

Block clear. Answer by 13. 
Block wanted. Answer by 2 or 5. 
Train has entered block. Answer by 13. 
Block is not clear. 
Train following. 

Opening block station. Answer by numbers of trains in 
the extended block with time each train entered the block. 

9. Closing block station. Answer by " 13 " after receiving 
transfer of the records of trains which are in the extended block. 
13. I understand. 

71. Train following, display " Stop " signal. Answer by ii 
S D. Stop signal is displayed. 

3 i7A. Rule 3i;A is for absolute block for following and opposing 
movements on the same track. 

To admit a train to a block the signalman must examine the block 
record, and if the block is clear, will give " i for No. 2 " to the next 
block station in advance. The signalman receiving this signal, it 
block is clear, must display the " Stop " signal to opposing trains, and 
reply " S D for No. 2." 'if the block is not clear, he must reply ' 5 
of No. 7." The signalman at the entrance of the block must then 
display the proper signal indication to the train to be admitted. 

A 'train must not be admitted to a block unless it is clear, except 
as provided in Rule 331 or by special order. 

3178. Rule 3178 is for absolute block for opposing moven 
and permissive block for following movements on the same track. 
To admit a train to a block the signalman must examine the 

* In The Railway Age of Chicago, I7th May, 1907, it was stated 
that 34,493 miles of single and 8,357 lines of double (or more) lines are 
protected by " manual-telegraph " 'block and 1,088 miles of single and 
Si 7 miles of double (or more) lines by " manual controlled " or " Lock- 





record, and if the block is clear, will give " i for No. i " to the next 
block station in advance. The signalman receiving this signal, if the 
block is clear, must display the " stop " signal to opposing trains .and 
reply " S I) for No. i." If the block is not clear, he must reply " 5 
of No. 4." The signalman at the entrance of the block must then 
display the proper signal indication to the train to be admitted. 

A train must not be admitted to a block which is occupied by a 
passenger train, except as provided in Rule 331 or by special order. 

To permit a train to follow a freight train into a block, the signal- 
man must give " 71 for Ex. 195 East " to the next block station in 
advance, to which the reply "5 of Ex. 187 East, S D for Ex. 195 
East " must be made. The approaching train will then be admitted 
to the block under " Caution " signal or with caution card (Form B). 

3iS.\. Rule 3i8A is for absolute block for following movements 


To admit a train to a block the signalman must examine the block 
record, and if the block is clear will display the proper signal indication 
to the train to be admitted, reporting its movement as per Rule 319. 

A train must not be admitted to a block unless it is clear, except 
as provided in Rule 331 or by special order. 

3i8B. Rule 3188 is for permissive block for following movements 

To admit a train to a block the signalman must examine the block 
record, and if the block is clear will display the proper signal indication 
to the train to be admitted, reporting its movement as per Rule 319. 

A train must not be admitted to a block which is occupied by a 
passenger train, except as provided in Rule 331 or by special order. 

A train may be permitted to follow a freight train in'.o a block 
under " Caution " signal or with caution card (Form B). 

319. When a train enters a block the signalman must report the 
train and the time to the next block station in advance, and when the 
train has passed the home block signal and the signalman has seen the 
markers, he must display tho " Stop " signal, and when the rear of 
the train has passed 300 feet beyond the home block signal he must 
report the train and the time to the next block station in the rear. 

This information must be entered on the block records. 

320. Unless otherwise provided, signalmen must not give " i " 
or '' 3 " until they have received " 4 " from the block station in the 

325. When a train takes a siding the signalman must know that 
it is clear of the block before giving " 2 " or displaying a " Clear " 
signal for that block. 

The signalman must obtain control of the block before permitting 
trains at starting and junction points, or on sidings, to enter the block. 

326. To permit a train to cross over or return, the signalman 
must examine the block record, and if all the blocks affected are clear 
of approaching trains he will arrange wilh the signalman at the next 
block station on either side to protect the movement, and when the 
proper signals have been displayed permission may be given. Until the 
block is dear no train must be admitlrd in the direction of the cross- 
over switches except under " Caution " signal or with caution card 
(Form B). All cross-over movements must be entered on the block 




Block Station M I9--- 

To Conductor inn! Eiigineman : Train No on track. 

Block is not clear. You may proceed with caution, expecting 
* to find track obstructed. 

Si^ii' [man. 

Conductors and enginemen receiving this card properly filled 
out and signed by the signalman may proceed with the train 
under control, prepared to stop short of any obstruction in the 


331. If from the failure of telegraph line or other cause a signal- 
man be unable to communicate with the next block station in advance, 
he must stop every train approaching in that direction. Should no 
cause for detaining the train be known, it may then be permitted to 
proceed, using a " Caution " signal or a caution card (Form D). 



.Block Station M. 


To Conductor ami Etigincman: Train No on track. 

Bell circuits and telegraph line have failed. You may proceed 
with caution, expecting to find track obstructed. 


Conductors and enginemen receiving this card properly filled 
out and signed by the signalman may proceed with the train 
under control, prepared to stop short of any obstruction in the 



The Chesapeake and Ohio R.R. has long been regarded 
as one of the best operated lines in America, "and reference is 
made to this in Chapter X. By the courtesy of Mr. Doyle, 
the general manager, the Author is enabled to give the block 
code in use on that railway : 

301. Block signals, unless otherwise provided do not affect the 
rights of trains under the time-table or train rules ; nor dispense with 
the use of the observance of other signals whenever and wherever they 
may be required. 

302. The normal indication of home block signals is " stop." 

306. A register is required at each block station. It must be ex- 
amined before a train is admitted to a block. 

A train passed by another train at a block station must be re-entered 
upon the register. 

The last train leaving must be the last recorded. 

307. When notice is received of an approaching train, the signal- 
man receiving it will notify the signalman in advance, ascertain if the 
block is clear and the " stop " signal displayed (using signal " 17 " 
for westbound trains and signal " 18 " for eastbound trains); and 
after arranging with that signalman for the train, display " clear " 
signal to allow it to enter the block. 

When a train enters a block the signalman will report it to the 
signalman in advance, and when the rear of a train has passed the 
block signal and the markers are seen, the signalman will display 
signal in " stop " position, and when the train has passed 300 feet 
within the block, report to the signalman in the rear that the train is 
clear of that block. When the block is clear, the home (and distant) 
signal shall be cleared sufficiently in advance of approaching trains to 
avoid delay. 

308. If a train takes the siding at a block station, the signalman 
must know that it is clear of main track before reporting block clear 
or giving " clear " signal to a train moving in opposite direction. 

311. When it is necessary to allow a train to follow another into a 
block, the signalman will issue caution card (Form B) when authorised 
by the train" dispatcher or by Rule 319. A train shall not be allowed 
to enter a block occupied by a passenger train, except as provided in 
Rule 319 or by special order. 

314. When necessary for a train to cross over at a block station 
where there are two running tracks, the signalman must first ascertain 
if the block to the rear on the opposite track is clear ; if so, ^arrange 
with signalman at the entrance of that block to display " stop " signal 
before allowing movement to be made. 

316. When necessary for a train that has been reported " clear ' 
to back into a block, the signalman must first ascertain if block is clear 
before allowing movement to be made. 

319. If from failure of telegraph line or other cause a signalman 
is unable to communicate with the next block station, he will stop 
trains moving in that direction, give to each written notice of the 
trouble, deliver caution cards allowing a train to follow ten minutes 
behind a freight train and twenty minutes behind a passenger train. 

331. Trains may pass a block station which displays a " clear " 
signal, but must not pass a " stop " signal unless caution card (Form 
B) is received, or it is necessary to do so in order to clear the mam 
track for a superior train. 

336. A train having cleared a block must not back into or within 
300 feet of such block without permission of the signalman. 

337. At block stations where there are two running tracks a tram 
must not cross over without permission of signalman, in addition to 
protecting train as per Rule 99. 

340. At a block station where the signalman is absent or in- 
capacitated, so that instructions cannot be obtained, a train shall wait 
ten minutes and then proceed with caution to the next block station, 
where the conductor must report accordingly to the train dispatcher. 

343. When approaching a block station, the engineman and fire- 
man will announce to each other, the indication of the signals. 

Rule 99, referred to in Rule 337, reads as follows :,-- 
99. When a train stops or is delayed, under circumstances in 
which it may be overtaken by another, the flagman must go back 
immediately with stop signals 'a sufficient distance to insure full pro- 
tection. When recalled, he may return to his train, first placing two 
torpedoes on the rail when the conditions require it. 

The front of the train must be protected in the same way, when 
necessary, by the front brakeman or fireman. 


There is nothing of recent years that has assisted a signal- 
man more in his work than the use of the telephone. On most 
lines nearly every signal-box is equipped, and some of them 
are on circuits between adjacent signal-boxes. By the use of 
telephones the men arc not only able to communicate quickly 
and frequently with their neighbours, but the running of 
trains is promptly and freely reported, whilst their use in 
case of mishap is very valuable. 





Where a box does not form part of a telephone circuit, 
and communication need only be given to the next box, there 
is no necessity to run a separate telephone wire, as it can be 
put on the block bell wire, and a special bell code can be 
arranged to call the man up to the telephone. 

Morse Code Sttective Instrument. 

The placing of several boxes or offices on one telephone 
circuit is often a nuisance, as the attention of a busy man is 
detracted from his duties to listen to the call to see if it be 
his. This may be avoided by the adoption of the selective 
telegraph instrument of the Morse Code Signal Co., of Mil- 
waukee, YVis. 

Each instrument is provided with what may be regarded 
as a Vale key, and has a wheel with spokes in. These are 
driven in by a hammer that makes long and short strokes 
corresponding with the dots and dashes of the Morse code. 
The Vale key rests on these spokes, and when the spokes 
have been driven in that correspond with the wards in the 
key, the latter falls and causes the circuit to be completed to 
the telephone. 

Supposing, then, that six signal boxes A B C D E F 
are on one wire, and A wants to speak to E, whose call is 
three short beats and two long. Instead of the bells in 
B C D E F all ringing none would ring, but the selective 
instruments would quietly work until the one in E corres- 
ponded to its key, when the bell would ring. 

The idea is also used for other purposes. 
Switching-out A ppa ratus. 

One of the necessary instruments for some signal-boxes 
is a switch whereby the block instruments can be switched 
out of use. There are certain times when the traffic con- 
ditions do not require a box to be 
open, and it may be closed 
although trains are still running. 
It may be at a private siding where 
no trains call at night, or a junc- 
tion where the branch traffic ceases 
after a certain hour. In order 
then to allow the box to be closed, 
the block wires have to be switched 
through from A to C in order that 
B may be closed, and therefore the 
signals that A and C exchange with 
B have to be exchanged between 
the first two named. 

A switch instrument such as is fixed in a box to be closed 
is illustrated in fig. 27. The switch bar a has three springs 
b l , b-, b 3 , which, when the box is open, complete the circuit 

between d l , d-, e 1 , e 2 , /*, f*, and lead to the instruments in the 
signal box, but when the box is closed, the switch is moved 
over and completes the circuit between d 3 , d*, e 3 , e*, f 3 , / 4 , and 
passes on to the next box. 

Fig 273 shows the wiring leading through the switch to 

Kig. 2711. Connections for Switching-out Apparatus. 

the instrument. 

The switch may be increased in length according to the 
number of instruments, and in fig. 27 a five bar switch is 

Controlling Switching-Out Apparatus. 

In order that an intermediate signal box may not be 
switched in or out without the knowledge and consent of the 
signalmen on either side the Sykes Co. control the switch in 
the manner shown by fig. 28. 

The intermediate box is represented by B and is shown 
in the diagram " in." To close the box the signalman at B 
lowers his stop signals and then the distant signals operated 
B c 


Fig. 27. Switching-out Apparatus. 

28. Controlling Switching-Out Apparatus. 

by the levers a a. He then rings A and C as per code and 
pushes in the plunger P. The men at A and C then hold 
over the switches b b and press down their bell keys which 
cause a current to enter the switch box c and to take locks 
out of the rod d which the man at B then draws to the right 
and indicates to himself " out " and back-locks the distant 
signal levers in their " over " position. At the same time 
the electric locks in the switch box c enter other slots in the 
"out " position of the rod d, and these have to be similarly 
withdrawn when the box is to be switched "in." The 
indicator e shows " locked " and " free " according to the 
condition of the lock from A and e- represents those from C- 



Light Indicator. 

In order that a signalman may know that the lights are 
burning properly in those signals of which he cannot see 
cither the front- or back-lights, it is customary for most 
British companies to provide a recording instrument for such 
signals whereby the light is electrically repeated in the signal 
box. This is not, however, done on all railways ; for instance, 
the Lancashire and Yorkshire R. and the North Eiastern R. 
are among the dissentients. 

Whilst there are several types, the general principle of 
light indicators is about the same, so it will be sufficient to 
show one form which is illustrated by tig. 29. This is the 
Skes indicator. 

f "t 


Fig. 29. Light Indicator. 

In the signal lamp there are two pillars, between the top 
of which, and over the flame, is suspended a copper bar E. 
When the lamp is lighted, this bar expands with the heat and 
the ends, when sufficiently expanded, cause an electrical cir- 
cuit to be completed and the indicator B in the signal box to 
show " in." Should the light get low or go out, the bar K 
contracts, the circuit is broken, the indicator turns to " out 
and the armature of the relay A falling, causes the bell D to 
ring. The ringing continues until the signalman turns the 
switch C. 

On the Great Western R., where a number of signals that 
are close together are provided with light indicators, it is only 
customary to fix one common instrument in the signal box. 
This shows " in " when all the lights are burning, but should 
one go out the instrument shows " out " and the bell rings. 

This is an economy in wire and instruments to which no 
exception can be taken, as it is immaterial to the man which 
light is out. 

Light indicator instruments should, as far as possible, be 
placed over the levers operating the signals they apply to. 

Electrical Repeaters. 

In the same way that those lights that cannot be seen by 
the signalman are repeated into the signal box, should the 
state of those signal arms be repeated that the signalman 
cannot see. 

Such instruments are known as electrical repeaters and 
they record when the signal is "on" and "off." A third indi- 
cation is generally given by a mid-position which intimates 
that the arm has not responded to the lever and come " off " 
or gone " on " fully. 

All distant signals should be repeated. This course is 

advocated owing to the importance of the signal and of the 
distance it is from the lever working it. 

Some companies attach the commutator of the instrument 
to the signal upright-rod and not to the signal arm, but as it 
is the state of the arm that is required to be known it is essen- 
tial that the commutator be attached to the arm. 

Where signals are slotted and consequently two or more 
men arc interested and each man has to be certain that his 
portion has worked properly the balance weight operated by 
each is the part to be repeated. Here, therefore, should be 
considered the question whether the state of both the weight 
and the arm should not be electrically repeated, as is the prac- 
tice on the North London R. (fig. 36). 

Most companies, in the case of slotted signals, do not pro- 
vide repeaters where the signal-arm concerned can be seen by 
the signalmen interested. This is wrong, as the slotting appa- 
ratus may have failed and yet the arm be fully "on " owing 
to the action of the other man. 

The Author considers that a 
generous use of electrical repeaters 
is money well spent. 

A signal may be regarded as 
"on " when the arm is between 5 
degrees above and 5 degrees below 
the horizontal line, as shown in the 
diagram in fig. 2gu. 

Where arms clear to an angle of 
60 degrees the " off " position may 
be as in the diagram, viz., between 
55 and 80 degrees. 


Fig. 2QA. 

Repeater instruments vary in shape. That repeater used 
on the L. and South \Yestern R. is shown by fig. 30. When 
at danger, contact is made between the spring a and the con- 
tact maker />, so indicating " on." When the signal is " off," 


Fig. 30. Repeater, London anil 
South Western Railway. 



Fig. 31. Mercurial Electrical Repeater. 

the insulated piece c forces the spring a from b into contact 
with spring d, causing indication " off " to be given. Should 
the signal arm not travel the proper distance, the spring a 
would be separated from both b and d, and cause the indica- 
tion " wrong " to be given. 

Another pattern of repeater is illustrated in fig. 31. This 
is attached to the signal arm, and consists cf a cup contain- 
ing mercury. When the arm is at danger the mercury is at 
the bottom of the cup and completes the circuit, showing 
"on." When the arm is down, the mercury flows to the 
other end of the cup and completes a circuit, showing " off." 
In case the arm does not come fully " off " or go properly 
"on," the mercury could not complete either circuit and 
" wrong " would be indicated. 

Fig. 32. Tver's Circular Repeater. 

Fig. 32 represents Tyer's circular shape. It is fixed on the 
front of the block instrument shelf and consists of a metal 
case and is 4|in. diam. so that where two or more levers that 
are provided with repeater instruments come together in the 
locking-frame there is space for each to be over the lever it 
applies to. 

Fig. 33 illustrates Tyer's box repeater. Both these instru- 
ments show three positions by means of the miniature arm : 
" on," " off " and a mid-position, indicated by the arrow, 
which signifies that the arm has not either gone fully to the 
"on " or the " off " position. 



TYE R'S Patent 

riR- 34- 

Tver's Repeater 
and Light 

33- Tv-er's Box Repeater. 

Fig. 34 shows an instrument in which there is combined 
both a repeater and light indicator. 

Fig. 340 shows the combined repeater and light indicator 
instrument used by the L. and North Western R. 

The lower of the illustrations shows on the right the in- 
strument, the repeater taking the form of a miniature arm. In 
the upper is a back view with the case removed. Suspended on a 
shaft are two balanced armatures a a-, and on the same shaft 
the miniature arm b turns. Above the armatures are two coils 
c c 2 , and above these a permanent magnet d. When the 
signal is properly " off " a current, switched in by a contact 
on the signal arm, enters the coil c, attracting the armature 
a and causing the miniature arm to indicate " off " as shown 
in the illustration. When the signal arm goes to the 
"on" position, the coil c is de-energised and coil <? 
excited so that the other armature a- is attracted, and this 
raises the miniature arm. In case neither coil is excited, 
owing to the signal arm not coming off properly or not going 
fully on, the disc e remains before the opening and indicates 
" wrong. " 

For the light indicator a permanent indication " out " is 
shown at the opening /, but when the expansion bar in the 
signal lamp makes contact the coils g are excited so that the 
disc h is raised behind the opening / and indicates " in." 

A new form of mercurial repeater has been introduced 
during recent years on the South Eastern and Chatham, the 
Great Northern, the L., Brighton and South Coast and the 
Lancashire and Yorkshire railways. This is the Pearson 
repeater shown by fig. 35, and which has been introduced by 
the Pearson Fire Alarm Svndicate. Westminster. It is recog- 
nised that one of the troubles of mercurial contacts is due to 
the oxidisation of the metal. The Pearson contact overcomes 
this by fusing the platinum wires into a sealed glass tube con- 



Fig. 340. Repeater and Light Indicator, I_. and N.W.R. 

taining a small quantity of mercury, and by adding a small 
quantity of special rich liquid hydro-carbon, the mercury and 
platinum remain perfectly clean. 

Fig. 35. Pearson's Repeater. 

Repeaters for Controlled Signals. 

It is necessary for the signalman, whose signal is con- 
trolled or slotted from another box, to know the position 
of the slot, so that should he work his lever and the signal 
does not respond he will know whether this be not due to the slot 
being on. This may be done by a mechanical disc in his box 
coupled to the slot on the signal ; but as this adds weight to 
the portion worked from the other box, and also introduces 
complications owing to difficulties in adjusting the wire, it is 
desirable to provide an electrical indication instead. This may 
be of similar form to an electrical repeater, but the instrument 
should be lettered " slot-on " and "slot-off." 

TVER'S Patent 


Fig. 36. Pryce and Tver's Slot Repeater. 

Another difficulty associated with slotted signals is, that 
each signalman should know the position of both slot and 
arm. It is necessary for each man to know the position of the 
slot the near man for the sake of knowing whether the slot 
be " off " or " on " and the far man to know whether his 
slot be working. It is also equally important that they should 
each know that the arm has gone to danger (in case they 
cannot see it). Therefore, it seems desirable that a double 
record should be given, which ma}' be on separate instruments 
or on a combined instrument, as shown in fig. 36. 

This is the Pryce and Tver slot repeater, and the upper 
indicator shows the position of the arm by means of a commu- 
tator on the arm as shown or by any other style of commu- 
tator, and the lower is the position of the slot given by means 
of a commutator attached to the slot lever. 

In fig. 37 are given details of some of the forms of repeater 

Small Indicator. Zinc Cover ' ~VerlicaL Needle Type 


Vertical Needle 

w IT 




Small Indicators for Tronl Board 4"CenIres 

3 positions 

Water hg\l double caiesfi 

Fig. 37' Forms of Repeater Instruments. 



Fig. 38. Electrical Repeater Indicalors and 
instruments supplied by W. R. Sykes Interlocking Signal Co. 

Combined Instruments. 

Repeaters, light indicators, slot repeaters, &c., may all be 
on separate instruments as has been said, but as it is most 
important that all instruments applicable to one signal should 
be fixed as near as possible to the lever working the signal, 
it is desirable to combine the various indications in one instru- 
ment. This is done very effectually on the Great Western R., 
and the Author is able to indicate by fig. 38 the various forms 
of instruments used on that line. 

Train Describer. 

For warning a signalman or the station staff the class, 
origin or destination of a train that is approaching the tele- 
phone is very useful, but it is still better to provide an instru- 

revolve so as to show a similar signal. When the pin is 
released the commutator completes the round, and rests in 
position for the next signal. 

A modification of this instrument has been made by .Mr. 
David Wells, of the North British R. Telegraph Dept. 

It is shown in fig. 40 and has for its object the advising of 
the station staff and waiting passengers at what platform 












Fig. 39. Tver's Train Describer. 

ment, as then the signal is on record and can be seen by all 

In Tver's Train Describer, fig. 39, A is the receiving dial 
showing what train is coming, B is the sending portion, with 
holes in which the pin D is placed opposite the signal to be 
sent. On the button E being pulled out as far as F, the com- 
mutator in the instrument revolves as far as the pin will allow, 
which causes the needle in the corresponding instrument to 

Fig. 40. Platform Indicator. 

certain trains will arrive. The instrument seen on the left is 
fixed on the platform and the other instrument is fixed in the 
signal box, where the signalman first announces the train on 
the upper dial and then the platform on the lower dial. Two 
line wires and 12 No. 2 Leclanche cells are required. 

At the Grand Central Station, New York, the arrival of 
trains is announced from the signal box outside to the station 
staff by means of the Teutelograph. The description of the 
train and the road it will arrive in are written on a record 
by the operator, which is reproduced on the instrument in the 
station several hundred yards away. 

It would be opportune here to refer to the Tclcmagna- 
phone that has been installed in the waiting rooms of the 



same station by the Callaphonc Co., of 51, West Thirteenth 
Street, New York. By this invention notice of the departure 
of trains is made into a transmitter, having five finely ad- 
justed microphones, and this is loudly repeated simultaneously 
from any number of announcing instruments, which may 
be placed in the waiting rooms, booking halls, etc. 

Communication for Controlled Level-Crossings. 

Where the gates at a road level-crossing are controlled 
from a signal-box some distance away, as in fig. 41, and the 
signalman wishes the attendant to close the gates across the 


Siynaf Box 

j Gafsmans 

Fig. 41 

roadway and open them for railway traffic, the arrangement 
of McKenzie & Holland might, with advantage, be applied. 

This is illustrated by fig. 42. Behind the gate-lever in 
the locking frame is a pillar containing a bell plunger b and 
an indicator c. On the gate post at the crossing is a bell d. 
When the signalman wishes the gates to be closed against 

Fig. 42. Controlled Level Crossings. 

the road he presses in the plunger b, which 
starts the bell d ringing, and it con- 
tinues ringing as the plunger is locked in. 
When the gates are closed across the roadway an electrical 
contact is made at e which causes the indicator c to point to 
" railway " and the signalman may then reverse his lever 
and secure the gates and this frees the plunger b so that it 
can be pulled out and the bell cease ringing. 

Electric Alarum for Level-Crossings. 

The object of the electric alarum for level-crossings is to 
warn the drivers of road vehicles that the gate-stops in the 
roadway are about to be raised and the gates closed. 

In McKenzie & Holland's arrangement an electric bell is 

fixed near the crossing which commences ringing as the 
gate-stop lever is pulled over to raise the gate-stops in the 
roadway. The bell is caused to ring by an electrical contact 
on the lever which is completed when the lever is partly over 
and broken when the lever is fully over and the gates are 
across the roadway and locked. 

Indicators for Gate-houses. 

At level-crossings that are not controlled from signal-boxes 
but by gatemen residing in adjacent gate-houses indicators 
working on the block-wire shculd be provided, so that the block 
signals passing between the signal-boxes on either side of the 
crossing may be recorded. The gate-keeper can then see 
whether any trains are approaching and whether they have 
passed the signal-box in the rear. 

Warning Bells jor Level Crossings. 

For level crossings where gate-keepers are not employed, 
a warning bell is desirable, which will commence ringing 
when a train is approaching and continue ringing until it 
has passed. 

Messrs. Stevens & Sons, the Sykes Co. (Lopes' patent), 
the Hall Signal Company and others have methods by which 
this is effectually done. 

At the crossing is fixed a bell with connections to an 
electrical treadle (fixed the required distance a\\av), which 
opens a relay, and causes the bell to sound until the train 
goes over another treadle at the crossing, which reverses 
the relay and stops the bell. Or it may be combined with a 
Track-circuit of the required length, which will cause the bell 
to ring when any part of a train is on the Track-circuit. 

Figs. 43 and 44 show the method adopted by the Railroad 
Supply Co., Chicago, of doing this on single and double lines. 

-1000 TO 3000 F+. 1 i-* WOO t O 3OO Ft. --4 


I < 


Fig. 43- 

A track-battery a is employed in connection with a section 
of " Track-Circuit " between the insulated joints g g and con- 
nected by the wires c to the magnets e. As soon as a train 
enters on the section the magnet e is de-energised and the 
armature falls away and joins up the bell / to the local battery b. 

f /OOff fo 300O Ff. 1000 To 3OOJ Ft *f 

Fig. 44- 


Bells for Warning Shunters, &>c., of Approaching Trains. 

Many accidents have happened to shunters, goods guards, 
etc., owing to their stepping on to a running line (to, perhaps, 
signal to their driver) when a train has been approaching on 
the running line of which they have been unaware. This 
may be due to a curve in the' road, station buildings, bridges 
or tunnels, or possibly to the noise of the approaching train 
being drowned by other noises. 

Such a warning bell as those just described should be 
provided at all points where shunters, etc., have to step on 
to a running line to signal to the drivers, and of which run- 
ning lines there may be an indifferent view. 

Trains Standing at Signals during Fog. 

Mr. Henry Jackson, the electrical engineer of the Lan- 
cashire and Yorkshire R., designed a useful little arrange- 
ment which has been a great boon to enginemen and signal- 
men during fog. 

When a train or light engine arrives at a stop signal dur- 
ing fog, and the signalman cannot see the engine or know 
that it has arrived, it may stand there for some time. 
Whistling is not always effective, as in busy places other en- 
gines are about. 

The arrangement under notice is the provision on the 
signal-post of a plunger which the fireman depresses on 
arrival, and that causes a signal to be shown behind the home 
signal lever which indicates "train-waiting." When the 
signal lever is worked to lower the signal to admit the train 
the instrument is automatically restored. Tver and Co. , Ltd. , 
are the makers. 

Signalling Long Tunnels. 

Although not actually connected with signalling, mention 
may perhaps be made of the arrangements made by some 
companies to provide for the safe working of long tunnels, 
such as the Woodhead on the Gt. Central, the Severn and 
Box on the Gt. Western, and the Totley, Cowburn, and 
Disley tunnels in the Peak of Derbyshire on the Midland. 

The arrangements are made so that immediate notice may 
be given to the signalmen at each end of the tunnel of a train 
or engine coming to a stand in the tunnel, of an accident, or 
of the line being unsafe from any cause. 

The first application of such an apparatus was made in 
the Woodhead tunnel, or, to speak correctlv, tunnels. There 
is an up and a down tunnel which are connected by 25 man- 
holes or openings. In each of these manholes there is a 
single stroke bell and two plungers, a plunger for each road, 
the bell being common to both lines. In the signal boxes on 
either side there is a special block instrument for each line, 
four instruments in all. Should any of the plungers in the 
manholes be depressed, the bells in the signal boxes start 
ringing. The instruments for the line plunged for up or 
down have an indicator marked "accident," which appears 
on a screen and the needle of a dial points to " out-of-order." 
This latter indication relates to the instrument, and the 
needle remains pointing to "out-of-order" until the instru- 
ment is reset by the telegraph linesman. 

The following code of signals is used : 

Failure of engine Two beats. 

Engine or vehicle off road Four ,, 

Broken rail or other obstruction, 
including falling of rock on to per- 
manent way Six ,, 

Accident gang required ... ... Nine ,, 

Obstruction removed ... ... ... Six pause six. 

This code is fixed inside the door of each plunger case, and 
a code is also provided for indicating the manhole where the 
breakdown, &c., has occurred. 

The most recent work of this sort is the Disley tunnel on 
the Midland R. The arrangements there are different to 
those at Woodhead. A special wire runs through the tunnel, 
which, if cut or broken, causes bells to ring in the signal 
boxes on either side of the tunnel. When the bell com- 
mences ringing, the signalman must not allow anything to 
enter the tunnel until he has been assured that the line is 
clear. He must, when the bell rings, give the " obstruction- 
danger" signal to the box at the other end of the tunnel, 
unless there is a train already in the section. In the latter 
event he must not give the " obstruction-danger " signal, 
but telephone the other box as to the bell ringing. 

The special wire in Disley tunnel is fixed on the up side 
about six feet from the ground. On either side of each man- 
hole and midway between the manholes there is a loop in the 
wire which should be cut in case of any train or engine 
coming to a stand in the tunnel, or if the platelayers or any 
other servant of the company consider that the line is unsafe. 

Servants concerned must not rely solely on the cutting of 
the wire, but must arrange for the proper protection to be 
given as laid down in the company's regulations. 

When the wire is cut, the telegraph linesman must be 
sent for to repair the cut, and when this is done the bells 
must be tested as prescribed. 

" Lock-and-Block " is worked through Disley tunnel. 

Repeaters for Fogging Pits. 

Some companies provide mechanical or electrical repeaters 
to advise fogmen of the condition of the signals they are 
working to. The disadvantages of mechanical repeaters are 
that they add weight to the signals, that they require adjust- 
ing, and that, not being used for 9 or 10 months in the year, 
they are liable to get out of order. The electrical repeaters 
are objectionable owing to the dampness due to their position 
affecting them. 

By the use of repeaters immersed in kerosene this objec- 
tion has been met by the Sykes Signal Co. As fixed on the 
L. and South Western R. and other lines the cotter pin that 
acts as a staple for the lid of the wooden case will, when 
inserted in the top of the case, join up the circuit from the 
signals to the repeaters. See fig. 37, page 14. 

" Train trailing " Indicator. 

For those places where signalmen cannot see trains 
standing at, or approaching, their home signals the indicator 
made by Messrs. Sykes, and illustrated by fig. 440, is very 
useful. Fig. 44& is a diagram of the electrical connections. 




When an approaching train is about 50 yards from the 
home signal it depresses an electrical contact which causes 
the indicator in the box to show train-waiting and a bell to 
ring. When the lever is pulled over the bell is stopped, but 
the indicator remains until the lever is put to normal. If it 

be inconvenient to lower the signal for a time a push button 
is provided to stop the ringing of the bell. 

" Clipston " Key for Intermediate Sidings. 
For siding connections on double lines that lie between 
signal boxes and which are too far out to be protected by a 
release- or bolt-lock an intermediate signal-box and signals 
has generally to be provided. If the traffic be light an 
Annett's lock will suffice, but as Annett's key has to be 

Bell fatter? 



H -About SOyds- -- 

Fig 440. Sykes' " Train Waiting " Indicator and Electrical Connections. Fig. 

Button to Stop tinging 
of Bell. 





Fig. 44C. " Clipston " Key for Intermediate Sidings. 



returned to the controlling signal-box before a second train 
can travel over the line, there is necessarily some time lost. 

Such sidings are, however, satisfactorily controlled by 
the " Clipston " key so called because it was first used at 
Clipston siding manufactured by the Sykes Co. and illus- 
trated by fig. 44c. 

On the lever working the points is a lock as shown in 
the lower illustration which is unlocked by a key. In a 
'' sending instrument '' in the signal box in the rear are a 
number of keys, and in the " receiving instrument " in the 
box in advance are other keys. As soon as a key is taken out 
of the " sending instrument " the block instruments for the 


Detached Vehicle Indicator. 

Vehicles are sometimes left on the main line and forgotten 
by signalmen, and to prevent this Messrs. Sykes have intro- 
duced the instrument illustrated by fig. ^d. 

On the platform is a switch instrument a, in the signal- 
box a bell b, an indicator c, a lock d in the tappet d" of the 
lever working the signal for the line, and a magnet e. When 
the signal has to be lowered the signalman pushes in the 
button /, which if the switch be normal joins up the battery 
of the switch so that the magnet e is energised and the lock 
withdrawn. When a vehicle is detached and left on the main 
line, the shunter raises the button a- in the switch instrument, 
which causes the indication at a 3 to change from white to 
red, the bell b to ring, and the indicator c- to appear before 

Indicating Switch 



Detached Vehicle Indicator 

Fig. 44<J. Detached Vehicle Indicator. 

section are broken down, but its insertion as long as 
another key has not been taken out - - in the " receiving 
instrument," synchronises the block instruments and they 
may again be used. The possession of a key by the guard 
allows the siding to be opened, and as soon as the unlocked 
point !ever is moved the key is " back-locked," and it cannot 
be withdrawn until the point lever is restored. 

In the upper part of the illustration is shown a case used 
by the station-master at the advance end to send back a 
supply of the keys. This is sent as a " value-parcel," and 
on its receipt at the rear station the keys are put in the 
" sending instrument." For this purpose each station master 
is provided with a key to the instruments. 

the opening c 3 . At the same time the battery of the switch 
is cut out so that the magnet e cannot be energised. 

The signalman may acknowledge by pressing plunger /* 
and ringing bell g on the platform. When the vehicle is 
removed the button a 4 is pressed, which causes a 2 to fall and 
all is normal. 


Some railway companies, the G. Western in particular, 
have saved considerable expense by installing phonopore 
telephones which can be superimposed upon any existing 
railway speaking or signalling circuit and so obviate the 
expense of erecting additional line wires. Excellent results 
are daily being obtained over circuits So, and even 100 miles 
in length. 

c a 



THERE are eight systems of " Lock-and-Block " in use 
in Great Britain, viz. : 

Sykes', Spagnoletti's, Hodgson's, Langdon's, Tyer's, 
Evans', Blakey & O'Donnell's, and Pryce and Ferreira's. 

Sykes' system is used throughout the South-Eastern and 
Chatham R., the Hull and Barnslcy R., the Mersey R., the 
VVirral R., and the Liverpool Overhead R. The L. and 
South Western R., the Great Eastern R., and the L. 
Brighton and South Coast R. have large sections fitted with 
it, and the Caledonian R., the North British R., and the 
Glasgow and South Western R. have small installations. 
Altogether it is used on 25 railways in this country, and over 
12,000 levers are interlocked by it. 

Spagnoletti's system is fitted throughout the Metropolitan 
R. and also on a portion of the Great Western R. 

Hodgson's system has been installed on a section of the 
L. Brighton and South Coast R. 

Langdon's system is on four or five sections of the Mid- 
land R., principally for the protection of long tunnels. 

Evans' system has been laid down on several sections of 
the Great Central R., also on the Wirral R. 

Blakey and O'Donnell's has been installed on one or more 
sections of the Great Northern R. 

Tyer's system is on the Caledonian R. 

Pryce and Ferreira's system is on the North London R., 
and also on several lines in India. 

Wood's system is in Australia. 

McKensie and Holland's system is in India. 

Siemens Bros, have also a " Lock-and-Block " system. 

The Sykes system is in use on about eighty per cent, of 
the lines fitted with " Lock-and-Block " in this country. 

Mr. Sykes is not the original inventor of " Lock-and- 
Block," although the Author believes that it was that gen- 
tleman who first designated the interlocking of the fixed 
signals with the block instruments as " Lock-and-Block." 

In the paper read before the Institution of Electrical 
Engineers in January, 1897, by the late Mr. Hollins, of the 
Great Eastern R., that gentleman said that if absolute block 
working be assumed to be the only real block working, then 
Mr. Tyer has an early claim, for both in 1852 and in 1854 
Mr. Tyer devised and had in use a rail contact for auto- 
matically and electrically indicating the passage of a train 

over certain points. In 1869 that gentleman patented a com- 
plete system of what is now known as " Lock-and-Block," 
and he (Mr. Tver) proposed to do electrically what is now 
done mechanically, viz., interlock one lever with another in 
the locking frame, and by his electrical slot control the signal 
for leaving a section from the cabin in advance (Journal of 
the Proceedings of the Institution of Electric Engineers, 
Part 126, Vol. XXVI.) 

Mr. Hollins also tells how in 1870 Mr. (now Sir) W. H. 
Preece, associated, he believes, with the late Mr. Langdon, 
then of the L. and South Western R. and late of the Midland 
R., introduced an electric lock on the L. and South Western 
R., to be worked in connection with Preece's block, and how 
Mr. Sykes in 1872 introduced a rail contact to electrically 
manipulate railway signals, and in 1875 introduced the first 
complete system of electrical interlocking, making it neces- 
sary to have the concurrence of three stations A, B and C 
to get two trains into any one section. 

Mr. Hollins concludes his historical remarks by observing 
that about 1882 Mr. Spagnoletti introduced his electrical 
interlocking on the Metropolitan R. That gentleman had, 
however, seven years previously patented his system of clear- 
ing the road, or resetting the block instrument by a rail 

Such is the earliest history of " Lock-and-Block." Mr. 
Hodgson brought forward his system in 1877, and Mr. 
Langdon in 1882, whilst the Evans' and Blakey & 
O'Donnell's are of recent years. 

The general public hears of " Lock-and-Block " in most 
cases from the government inspectors' reports on railway 


At one time these references were very frequent; but no 
mention has been made of the need of " Lock-and-Block " 
since the accident on the Cheshire Lines near Manchester in 
November, 1899. 

It must not be assumed from this that "Lock-and-Block" 
has been universally adopted, as very little has been done 
during recent years except on the L. and South Western, 
the Great Eastern (Suburban) and on the Scotch lines. The 
main trunk lines L. and North Western, Great Western, 
Midland, North Eastern, Great Northern, etc., have no in- 
stallations except a few isolated cases. 

This small progress, and probably also the present silence 



of the Board of Trade Inspectors, are no doubt due to the fact 
that " Lock-and-Block "cannot be regarded as perfect. During 
the last 6 vears there have been at least four collisions two 
in London and two serious and fatal ones in Glasgow on 
lines where " Lock-and-Block " was in operation. 

" Track-circuit " whereby signals are controlled by the 
state of the section ahead may replace " Lock-and-Block," 
or might with advantage be added to it, as it would remove 
all the imperfections of " Lock-and-Block," as signals could 
not then be lowered if a train or any part of a train were 
in the section. 

Another difficulty is that on a busy line " Lock-and- 
Block " would compel absolute block working to be adopted 
for goods trains. This no doubt would mean greater safety, 
but it would play great havoc on lines that have a heavy 
goods and mineral traffic. At the present time if a goods 
train be shunting at a box under the protection of the home- 
signal or be standing at the starting or advance starting 
signal awaiting " line-clear " a second goods train may be 
accepted from the signal-box in the rear under clause 5 of 
the Standard Block Regulations, which is as follows : 

5. Section-Clear-but-Station-or-Junction-Blocked. 

When the line is clear to the home-signal, and it is necessary for a 
train to be allowed to approach cautiously in consequence of an obstruc- 
tion existing ahead of the home-signal, or from any other cause, the 
is-line-clear " signal must not be acknowledged in accordance with 
regulation 3, but the " seclion-clear-but-station-or-junction-blocked " 
signal must be given, and when this signal has been acknowledged, the 
block indicator must be placed to the " line-clear " position. The sig- 
nalman receiving this signal must (if the train has not already passed 
the home-signal towards the starting or advanced starting-signal) bring 
the train to a dead stand at the home-signal, and verbally instruct 
the driver that the section is clear, but the station or junction ahead is 
blocked. A green flag by day and a green light by night must at the 
same time be exhibited to the driver, and the necessary fixed signals 
lowered to give permission for the train to proceed. The " train- 
entcring-section " signal must then be given, and acknowledged, and 
the block indicator placed at " train-on-line." 

Where the home-signal is at such a distance from the signal box that 
it is not possible for the signalman to communicate verbally with the 
driver when the engine is standing at the home-signal, the signalman 
must, after bringing the train to a dead stand at the home-signal, lower 
it to allow the driver to draw up to his signal box, and must stop the 
train at the signal box by exhibiting a red flag by day and a red light 
by night. The driver must then be verbally instructed that the section 
is clear, but the station or junction ahead is blocked ; after which a 
green flag by day and a green light by night must be exhibited to the 
driver, and the necessary fixed signals lowered to give permission for 
the train to proceed. 

If a train is assisted by an engine in the rear, a green flag by day 
and a green light by night must also be exhibited to the driver of the 
engine in the rear of the train. 

Except where special instructions are issued to the contrary, when a 
train has passed the signal box and is brought to a stand at the starting- 
signal or the advanced starting-signal, the driver must understand that 
the lowering of the starting-signal or the advanced starting-signal is an 
indication that the line is only clear to the home-signal at the signal 
box in advance, and that he must regulate the speed of his train in the 
same way as if he had been verbally instructed to proceed under the 
" sectioii-t1ear-btit-station-or-junction-b}ocked " signal. 

With " Lock-and-Block " trains cannot be worked under 
the " section-clear-but-station-or-junction-blocked " rule (ex- 
cept in Evans' system, described in Chapter III). This 
would lead to delays, and this fact has possibly assisted in 
retarding the progress of what is undoubtedly a great safe- 

Its main object is to prevent irregular block working, 
and some figures as to collisions caused through such errors 
made by signalmen may be of interest and service. 

Taking a certain ten years, it appears that the Board of 

Trade held enquiries into 45 accidents in which the reports 
go to show that irregular block working had contributed 
wholly, or in part, to the accidents. 

These figures show how many accidents " Lock-and- 
Block " would, or should, have prevented. 

But these figures also show how magnificently well our 
signalmen do their work. When one remembers the hun- 
dreds of millions of trains that have been signalled from one 
signal-box to the other during a year, and that on an average 
only five of these were so incorrectly signalled as to lead to a 
passenger-train collision, it is really marvellous, and reflects 
the greatest credit on the men concerned. 

In connection with this, it must be observed that there an? 
many railway officials of very high standing who object to 
the very idea of " Lock-and-Block," on the ground, they 
say, that it tends to mar the morale of the men, and that 
it spoils their reliance upon themselves. 

This is an argument that has been used before in other 
directions when it has been proposed to make some change, 
and it must be admitted that there is some force in it, for 
instance : 

On the morning of November 19, 1896, the signalman 
at Tottenham North Junction on the Great Eastern R. let 
an engine and nine wagons out of a siding on to the up line, 
and whilst they were standing there waiting to be crossed on 
to the down line the signalman accepted a passenger train, 
lowered his signals for it, and a collision ensued. 

This section of the Great Eastern R. is fitted with " Lock- 

Col. Addison enquired into the causes of the accident, 
and he devoted the greater part of his report to observations 
as to the signalman's not using an essential part of the 
apparatus, and made the following remark, which bears upon 
the point now immediately under notice. ' The fact that a 
" signalman of Ellis' experience, bearing as he does a very 
" high character as a useful and reliable man, could accept a 
" train in entire forgetfulness of having placed a goods train 
"on the line only two minutes previously *. " 

At Park station on the Wirral Railway on the morning 
of February 15, 1897, a passenger train from New Brighton 
to Liverpool came into collision with some empty carriages 
which were standing at the platform on the up loop line. 
The passenger train generally runs on to this line, but on 
the morning in question some empty carriages had to be 
placed there for certain reasons, and the Liverpool train 
ought therefore to have been turned on the up main line. 

Colonel Yorke's report says the signalman " was fully 
" aware of the position of these empty coaches, which were 
" nearly opposite to, and in full view of the windows of the 
" signal cabin *****. 

" No doubt his forgetfulness was momentary; but having 
" regard to the fact that he had himself caused the empty 
" carriages to be placed on the loop line, and that they were 
" nearly opposite to his signal box, his failure to make sure 
" that the loop line was clear before he lowered the signal for 
" the passenger train indicates a degree of carelessness on 
" his part which it is difficult to excuse." 

But after all these are only two isolated cases, and whilst 


perhaps it may be true that " Lock-and-Block " tends to re- 
move some of the self-reliance from the men, yet on the other 
hand there are the 45 cases already quoted which the system 
would, or should, have avoided, besides the unknown narrow 
escapes from collision, and the goods train accidents which 
are not enquired into. 

It may also be interesting and instructive to note some of 
those cases where the Board of Trade inspecting officers 
have stated that the collision could not have occurred had 
" Lock-and-Block " been in use. 

By this recital it will also be readily seen what are the 
purposes of this system. 

On November i3th, 1890, a goods train passed through 
Primrose Hill tunnel, and then a second goods train followed. 
Through some misunderstanding the man at the box in the 
rear was led to believe that the block signal " train-out-of- 
section " sent for the first train was the signal for the second, 
and so he asked for and received permission to send a pas- 
senger train, and a collision ensued. By " Lock-and-Block " 
the block signals could not have been sent, nor the starting 
signal lowered for the second goods train until the first had 
arrived, nor, of course, could the passenger train have fol- 
lowed until the second goods train had passed through the 

The accident at Thirsk on November 2, 1892, will pro- 
bably be remembered. Signalrrfan Holmes accepted a goods 
train, and then fell asleep, being worn out owing to having 
had no rest owing to his child's illness. He was rudely 
awakened by having the up Scotch night train offered to him, 
which he accepted, although the goods train stood at his 
home signal. By "Lock-and-Block" the unfortunate man 
could not have accepted the express until the goods train 
had been disposed of. 

It is not generally known that the Board of Trade took 
the unusual course of sending a copy of Major Marindin's 
report on the Thirsk accident to every railway company, 
drawing attention to the recommendations of the Inspecting 

On March 3, 1897, at Eastleigh, L. and South Western 
R., some carriages were placed on the down Portsmouth 
line to await the arrival of another train to which they were 
to be attached. Whilst standing there the signalman low- 
ered the signals for a train to go to Portsmouth, and a col- 
lision ensued, which would not have been possible had a system 
of " Lock-and-Block " been in use. 

These are typical illustrations of three different objects 
attained by the system, but another and last quotation will 
show a further most important purpose achieved. 

This refers to the Barassie accident on the Glasgow and 
South Western R., on February 4, 1898. The signalman 
there accepted a passenger train which had to travel from 
Kilmarnock to Troon new station, and in so doing it would 
have to cross the line from Troon old station. After he had 
accepted the passenger train he also accepted a goods train 
from Troon old station. He lowered his signals for the 
passenger train, keeping, necessarily, those for the goods 
train at danger. Unfortunately the goods train over-ran the 

signals, and a serious collision occurred, which might have 
been avoided by "Lock-and-Block," as the acceptance and 
signalling forward of the passenger train towards Troon 
new station would have interlocked the block instrument 
from Troon old, and so prevented the acceptance of the goods 

The protection of trains crossed on to a facing line 
is a very important matter, which was brought prominently 
before the public by the Norton Fitzwarren accident in 
1890. It is an operation that is performed hundreds 
of times a day, and there are two important points in 
connection therewith. Firstly the signalman shall not 
"clear" the line from which the train is shunted until the 
whole of the train has got through the crossover road, and 
secondly, that the signalman shall " block-back " to the box 
in the rear for the line upon which the train is to be shunted. 
The former movement will usually take care of itself, 
as trains are generally being shunted in order to make way 
for other trains, and so the proper block signals come 
naturally, but a special movement is necessary to " block- 
back " on the other road, and it is imperatively important 
that the block instrument in the rear cabin should indicate 
that the section is blocked. If the signal " train-out-of- 
section " be not allowed to be given for a train until it is a 
quarter of a mile in advance of the box, it is of equal im- 
portance that the block instrument in the rear cabin should 
show " train-on-line " when a standing train is at the next 
box. Of course, the block instrument shows " train-on- 
line " in the ordinary course for a running train, because it 
has to pass the rear box and be sent forward, but the man 
at the rear box knows nothing about the shunted train, and 
so it becomes part of the signalman's duty to give the pre- 
scribed bell code to the box in the rear, and for the latter to 
acknowledge it and turn the needle of his block instrument 
to " train-on-line." All this does not come into the regular 
routine of a signalman's work and is something out of the 
common, and consequently liable to be neglected, and so if 
anything can be introduced whereby this operation is per- 
formed automatically, or the signalman be compelled to 
" block-back " before shunting a train, a very important and 
safe duty will be carried out without any doubt or risk. 

A most important feature in " Lock-and-Block " is the 
electrical contact. It prevents a signalman giving "line- 
clear " before the train is actually on its way and has passed 
over the contact. Without it a signalman could give such 
a signal before the train has really arrived. A very difficult 
problem in connection with these contacts is the selection of 
the positions in which to fix them to meet all the conditions 
of the traffic. For instance, there requires to be more space 
between the protecting signal and the treadle for a long goods 
train than is necessary for an engine and brake. If the treadle 
be fixed, say, 400 yards from the signal, to accommodate a 
goods train, it is an inconveniently long way out for a shorter 
train. Yet, if the treadle be fixed at a shorter distance, then 
it would be possible to clear a train before the tail of it had 
passed the signal. 

There are two other complements to the system. The one 


is the means for releasing a signalman should he get " locked 
up," and the other is an apparatus for mechanically putting 
the starting signal to danger without the signalman's inter- 
vention and so preventing the signal from being left " off." 

That the former is necessary will at once be understood 
when it be remembered that should a signalman have offered 
a train and then, for some reason or other, it does not go 
forward on its journey, something must be done to unlock 
his block instrument and restore matters to the status quo 
ante. Then again, at some stations, trains are made up and 
two trains are joined together that have come from different 
directions. But the arrival of the first train would lock up 
the block instrument at the station, and it would, under 
ordinary circumstances, remain locked until the train went 
into the next section and so the second train could not be 
admitted although the two trains had to be joined together. 
Another difficulty arose when one train was shunted into a 
siding for another more important train to pass. The in- 
strument in that case would be locked up until the contact 
had been made. It therefore became necessary to devise 
something to unlock the instrument in the same way as if the 
contact had been reached. And so a releasing key was 

If " Track-Circuits " were adopted in connection with 
" Lock-and-Block " the train itself when going forward into 
the advance section would lock the signal lever and the 
instrument too if the latter be still necessary. If the train 
did not go forward for some reason then the signalman would 
not be locked up, and so no releasing key would be re- 
quired. The same would apply if the train were shunted 
into a siding, as directlv it was " inside " and the main line 
clear the signal and instrument would be freed. 

There is no doubt the releasing key is a source of danger. 
In the four cases quoted of accidents on " Lock-and-Block " 
worked lines the releasing key was used, and had "Track- 
Circuits " been in operation the men concerned could not 
have freed themselves. See also p. 2(>. 

An idea as to the protection obtained by controlling the 
signals at one signal-box by a "Track-Circuit" extending 
up to the next box, may be obtained when it is stated that 
the Board of Trade reported on 12 accidents during 'the 18 
months ending January 1904, that would in all probability 
not have occurred had such a system been in operation. 

The 12 accidents are as follows : 
Aug. joth North British, Charing Cross : 

Wrong block instrument (Lock-and-Block) used, line 
cleared improperly and second train admitted into section. 
Sept. I2th G. W. & Mid., Bristol: 

Signalman overlooked that first train had stopped sud- 
denly, and lowered signals for second train. 
Sept. igth Gt. Western, \Vestbourne Park : 

Signalman overlooked light engine standing on up line 
and lowered up signals for another train. 

June 5th District, Westminster Bridge : 
June I7th Metropolitan, King's Cross : 

Error in block working. Lock-and-Block signals cleared 
in error and second train admitted into section. 
Sept. 25th Caledonian, Glasgow Central (Underground) : 

Lock-and-Block cancellation signal improperly used and 
second train admitted into section. 
Oct. 22nd Lancashire & Yorkshire, Sowerby Bridge : 

Signalman overlooked light engine standing on up line 
and lowered up signals for another train. 
Nov. 6th North British, Lochmill : 

Error in block working. Signals mis-read and line 
cleared and second train admitted when first was in section. 

Dec. 2 3 rd Hull & Barnsley, Hull : 

Goods train broke loose, line cleared on arrival of first 
portion and second train ran into broken loose portion. 
Dec. 24th North British, Stobcross : 

Error in block working. Two trains in one section. 
Dec. 3ist L. & N.\V. & G.W., Chester: 

Train stood foul and signals lowered for second train 
to cross. 
Jan. jth Midland, Chinley : 

Error in block working. Two trains in one section. 

" Track-Circuits " are used principally for automatic 
signals and where power plants are in operation. They are 
also used for " Lock-and-Block " purposes in America and 
on the Continent instead of treadles or other electrical con- 
tacts. Sections of "Track-Circuit" are admittedly better 
than electrical contacts. 

For instance, where a contact is in use the very first wheel 
deflects it, and sends the releasing current, although the 
greater part of the train may not have cleared the contact, 
but where a section of " Track-Circuit " is used any length 
of rail may serve, and by that means, unless the whole ol 
the train has passed out of the section, the releasing current 
cannot be sent. 

The system also meets that other difficulty named just 
now as associated with the use of electrical contacts, and 
that arises through the releasing contact having to serve 
for an engine and van as well as a long goods train, and 
therefore it only need be a short distance in advance of the 
starting signal for the shorter train to free the block in- 
strument, but that short distance is insufficient for a long 
train, and consequently the contact has to be fixed at the 
greatest possible distance so as to clear everything, and 
therefore the engine and van have unnecessarily to travel 
that much further. With a short section of "Track-Circuit" 
as an alternative to a contact all that is required is to in- 
sulate a section commencinp- on the signal-box side of the 
starting signal, and terminating a short distance on the other 
side of the signal, and consequently, be the train short or 
long, the whole of it must be on its way before the electrical 
releasing current can be sent. 

Another difficulty associated with contacts is, that when 
they are provided so that a train or engine standing at them 
shall lock certain signals, or a block-instrument, there is no 
guarantee that the train or engine will actually stand in the 
limited space provided. For instance, when a train standing 
at a signal is out of the signalman's sight, an electrical con- 
tact is sometimes fixed that will give an alarm to the signal- 
man, and the block instruments may also be interlocked 
therewith. It ma}- so happen that the driver stops short of 
this point, and so the necessarv security is not obtained, as 
is also the case in stations where engines or trains stand out 
of the signalman's sight, and electrical contacts are pro- 
vided so that the signal for entering the road occupied by 
the train or engine is locked under those conditions, as there 
also the driver may not stand on the contact and vet be in 
danger. But, by insulating a section of the line any length 
may be protected, and all these difficulties overcome. 

That there has not been a more general use of the system 
is not due to prejudice. Until recently the applications were 
far from satisfactory. 

There was the almost fatal objection that the wheels of 
nearly all passenger vehicles are fitted with centre discs made 


of teak wood, which, being a non-conductor, prevented the 
current being short-circuited properly on the insulated rail 
section. It is now, however, found that the general use of iron 
brake blocks provides some compensation for this, as the 
current passes from the rail to the tyre of the wheel and then 
to the brake-block and so becomes short-circuited. 

On the L. and South Western R. , where there are many 
miles of " Track Circuits," the Mansell wheels are bonded by 
two thin copper wires from the tyres to the metal centre 
bosses of the wheels. 

Then there has also been a fear that the climatic con- 
dition of our country, owing to its excessive moisture, pro- 
duces a bad insulation. If this were so it might possibly be 
overcome by the use of a better material in the road judging 
from some remarks made in a report presented to the Inter- 
national Railway Congress (London, 1895), by Messrs. A. 
W. Sullivan and F. A. Delano, on behalf of the American 
Railway Association, on the subject of the Development of 
automatic signalling in the United States. Mr. Sullivan, 
when dealing with the question of reliability, says : 

"It may also be of interest to note that there occurred only one case 
of grounding of track battery sufficient to cause relay to open. Until a 
comparatively recent date it has been thought necessary to provide a 
clean stone ballast, clear of rails, in order to prevent short circuiting of 
track battery through the ballast. In view of the fact that the ballast 
in use on that portion of this (Illinois Central) railway is mainly cinder, 
or cinder mixed with gravel, and that owing to the difficulty of draining 
the eight tracks, because of peculiar local conditions, it is frequently very 
wet, and often in contact with the rails, it would seem that the troubles 
likely to be experienced from this source have been much exaggerated." 

Mr. Robert Pitcairn, of the Pennsylvania R.R., in his 

Progress during recent years, just when headway was 
apparently being made, has recently been further checked by 
the discovery that a single vehicle such as a goods wagon- 
placed on a section of "Track-Circuit" requires such a 
different adjustment in the relays to what is necessary for 
general running purposes. 

Such a movement, i.e., the leaving of a single vehicle on 
a running line, unprotected and overlooked, and coinciden- 
tally the non-shunting of the relay, forms such a fractional 
part of those movements that can safely be protected by 
" Track-Circuits " that this objection may surely be re- 
garded as a negligible one. It is only fair, also, to say that 
careful tests made on the L. and South Western R. and on 
the Great Central R., under similar conditions to those 
that led to the discovery referred to above, failed to find the 
difficulty mentioned. 

Further, it may be said that as "Track-Circuits" are 
here considered as part of the complement of " Lock-and- 
Block " that they do not require for that purpose the same 
delicate relays and instruments as are required in Automatic 

Trains shunted from one main line to another and stand- 
ing on the facing road and trains standing at home and 
starting signals and on the main line waiting to be shunted 
across or into a siding can by " Track Circuit " be safely pro- 
tected, so rendering it unnecessary for train-men to be in 
the signal-box to act as a reminder to the signalman as laid 

6 S 


4 3 


Z 1 


Is, la. 


1 ' 



T 2 ' 





report to the same Congress on another phase of the auto- 
matic signalling question, makes the following note : 

An objection to the track circuit of a purely theoretical nature has 
frequently been put forward with much apparent confidence by advocates 
of other systems, which is based upon the assumption that the leakage of 
the electric current from one line of rails to the other in very wet weather 
may be sufficient to divert the current from the signal operating magnet, 
and thus prevent the clearing signal from being shown. No such diffi- 
culty has been met with in practice, a fact which is not difficult of ex- 
planation upon theoretical grounds. Actual measurements made by Mr. 
B. H. Mann, upon a section of track, showed that the conductivity re- 
sistance of 7,980 feet of rails was only 0-52 ohms, while the minimum 
insulation resistance of the same section when the ground was very 
wet was 2'5 ohms; conditions which afford ample margin for success- 
ful work. The bulk of the current necessarily takes a longer route 
through the rail conductor, in preference to the shorter route through 
the wet ties (sleepers) and earth, the difference in their respective con- 
ducting powers as shown above being roughly as 5 to I." 

Another prejudice arose from the discovery on the Great 
Eastern R. and the Great Western R. that engines could 
stand on a sanded rail and not shunt the relay. 

down in Rule 55. 

For instance, it would have avoided the collision at Tap- 
ton Junction, Chesterfield, on the Midland R. on September 
nth, 1906, when during a dense fog a light engine was over- 
looked by the signalman, \vho lowered his signals and 
allowed a train to run into the light engine. 

Another important use of "Track-Circuits" may be 
noted. At certain large terminal stations there is room for 
two trains to be unloaded at the same platform, and the 
custom generally is to provide for such roads a stop-signal 
of the usual pattern to be lowered when the full length of 
line is clear, a calling-on arm to be lowered when only part 
of the station is clear, as is the case when the first train is 
already in, and then a third form of signal is necessary 
when both trains aic in, and a light engine has to be ad- 



tnitted to 'draw the empty coaches out or for some other 
shunting operation. Or, as is generally the case at terminal 
stations, an incoming train becomes an outgoing one, and 
then some signal is necessary to allow the fresh engine to 
back on to the coaches, as, of course, the whole platform 
being full, neither the stop" signal nor the calling-on arm 
should be lowered. 

The difficulty here is to safely and efficiently guarantee 
the lowering of the correct signal, and this may be attained 
by an installation of " Track-Circuit," which may be carried 
so far as to allow all three signals to be coupled to one 
lever, and then when anything has to be admitted the sig- 
nalman pulls over his lever and the correct signal is 
' ! selected " according to the state of the road. If the plat- 
form be clear throughout, then the top arm is lowered. If 
only half be clear, then the calling-on arm is freed, but if 
the two trains have already nearly occupied the road, then 
the shunting arm will be " selected." 

Or it may be applied to through stations like Birming- 
ham (New Street), York, Manchester (Victoria), &c., where 
there are signal boxes in the station and at the entrances. 
If the road through the station be clear up to the starting- 
signal at the other end, then the man at the middle box can 
not only lower his stop-signal (generally fixed in the middle 
of the station), but his distant-signal as well, and the man 
at the outside box would then lower his distant too, thus 
indicating to an approaching driver the state of affairs in the 
station. If the line be clear up to, and a short distance past, 
the middle box home-signal, then the man at the middle box 
keeps his stop and distant at danger, and the outside box 

man lowers his own home-signal but keeps his distant " on." 
Then when the line is just sufficiently clear at the entrance 
to admit a short train, the man at the outside box keeps his 
home-signal and distant-signal at danger, and having pulled 
the train up he admits it by a calling-on arm. All these 
movements may be efficiently and correctly controlled by- 
sections of " Track-Circuit " running through the station. 

In addition to automatic and power signalling plants 
there are a few sections of " Track-Circuit " in this country. 

On the Great Northern R. there are four one notable 
case being in the tunnel outside King's Cross. On the 
South Eastern and Chatham there is one at St. Paul's; on 
the Lancashire and Yorkshire R. there are several as a 
result of successful trials at Manchester (Victoria) and Liver- 
pool (Exchange); there are installations through the two 
tunnels between which Nottingham Joint Station is situated 
and there are also some installations on the Midland R. at 
Child's Hill, Manningham and other places. 

An example of what can be accomplished by " Track* 
Circuits " is to be found at Guildford Station on the L. and 
South Western R. , which is protected by two signal-boxes 
the South box at the London end and the Yard box at the 
\Yoking end. As the signalmen in these boxes have a very 
indifferent view of the lines in the station it was decided to 
equip these lines with "Track-Circuits," as shown in fig. 
442, and the work was carried out by the British Pneumatic 
Railway Signal Co., Ltd. 

Rotary slots, similar to those described in fig. 125, have 
been fixed on the signals at each end of the station, with indi- 
cating discs on the same wires in the signal-boxes. 

Fig. 44f. Track Circuits at Guildford: London and SouthAVestern Railway. 



Protecting Converging Junctions. 

In Colonel Yorke's report to the Board of Trade on the 
collision at Thringley Junction, G.VV. R., in Jan., 1907, the 
inspector commented on the difficulties railway companies 
have in operating converging junctions where the signal-box 
in the rear on one or both lines is some distance away. In 
the case under notice the distance from Thringley to Melk- 
sham on the branch line is four miles. As the branch line 
joins the main Great Western line from Bristol, Bath, etc., 
at Thringley, it follows that there is a large amount of traffic- 
on the main line which, if absolute block junction working 
were in force, would have to be kept back at the box in the 
rear whilst branch trains were travelling these four miles or 
all branch traffic must be kept back at Melksham until there 
was a period of sufficient length between the times of the 
main-line trains for a branch train to cover the distance. To 
meet this difficulty the " warning-arrangement " was in 
force whereby, under the " section-clear-but-junction- 
biocked " rule of the block regulations, a train was pulled up 
at the box in the rear and the driver warned as to the state 
of the junction and then allowed to proceed. Since the Derby 
Junction (Birmingham) and Esholt Junction accidents the 
"warning-arrangement" has been considerably less used, 
but railway companies still find its use necessary, and the 
Great Western Co. submit that Thringley Junction is such a 
case. Colonel Yorke, however, points out two alternatives, 
but he is careful to say that he does " not make any recom- 
mendation on this subject, but the consideration of the com- 
pany might be invited to it." The first is the provision of 
an intermediate block-post or " a less satisfactory method " 
the erection of outer home signals at (say) 100 yards from 
the junction. 

This latter proposal should meet the difficulty. It is the 
method adopted by the L. and South Western R., and onlv 
costs the price of the additional signals. One objection to 
this proposal and it is granted that there arc some--is that 
a driver may overrun the outer home and foul the junction, 
and another objection is that the signalman must then accept 
conflicting trains on his block-instruments. The intermediate 
block-post scheme is the more perfect, but look at the cost ! 
The outlay for providing the box, signals, block instruments, 
etc., will not be much under ^300, whilst the cost of man- 
ning the box with two men and the maintenance of the sig- 
nals, etc., would be about 150 per annum. 

The Author suggests the following arrangement, which 
would not only meet the case of intermediate block-posts 
outside junctions, but where they are necessary to break long 
block sections. The idea is to provide an advance starting 
signal for the box in the rear and which should be placed 
sufficiently far towards the next box to considerably shorten 
the intervening length and reduce the time the movement of 
one converging train holds back the other. This advance 
starting signal should be worked by power, with a line wire 

to the signal-box, where the full movement of the lever or 

of a slide or switch shall complete an electrical circuit to 
the signal. The signal could then be any distance from the 

box and " Track-Circuits " would be employed, as it is 
assumed that the signalman would not be able to see a train 
standing at the signal or know when it had passed. 

Assuming that the lines, etc., at Melksham and Thringley 
are as shown in fig. 44/, the suggested advance starting sig- 
nal 3 might be i, i^ or even 2 miles from Melksham. When 
a train could not be accepted by Thringley the man at Melk- 
sham would lower his starting signal i and allow the train to 
go forward to signal 3. Whilst on its way possibly the train 
would be accepted, and then he would lower signal 3. It 
would probably be an advantage to have a repeater signal 2 
of the distant type, which would be lowered through a circuit- 
breaker on signal 3. A " Track-Circuit " would be laid in 
from signal i to 400 yards past signal 3 so that signal i could 
not be lowered unless that section were clear, and an indica- 
tion would be provided in Melksham box to show the state- 

Fig. 44/. Protection of 
Converging Junction. 

Starting Signal for Melksham. 

Kepeatt-r Signal for No. 8. 

Advance Starting Signal for Me'.ksham. 

4. Distant Signal for Thringley. 

5. Home Signal for Thringley. 

of the section. A second " Track-Circuit " would be desir- 
able, although not absolutely necessary, from signal 3 to sig- 
nal 5, so as to control signal 3. Signals 2, 3 would be elec- 
trically repeated into Melksham box, and they would be pro- 
vided with replacers for automatically putting the signals to 
the " on " position. 

This arrangement would cost about ..200 to provide, and 
the annual upkeep would be about ,.50. A further advantage 
is that the signals would be always on duty without extra 

Something similar to this suggestion has since been 
adopted by the G. Western Co. between Pangbourne and 
Goring, of which details are given in the Appendix B, p. 336. 

Controlling Releasing Key. 

On p. 23 the possible dangers from the improper use of 
the releasing key were referred to. The Sykes Co. prevent 
this by fitting a shutter over the key-hole in the instrument 
and which can only be removed in order to allow the key 
to be used by the joint action of the signalmen on either side 
sending a releasing current simultaneously, and they in turn 
can only send the current after re-setting their instruments 
to normal, providing, of course, that the block-sections are 
unoccupied. The key being used must be taken out of the 
instrument at once to re-establish communication. Thus it 
will be seen that three men are required to act in concert in 
order to cancel one train, so making- risk of collision im- 



Sykes' System. 

THIS system is diagrammatic-ally illustrated by fig. 4^. 

Let it be assumed that there are three signal boxes, A, 
B. C, on a double line. The following would be the working 
on the up line. 

In each of the boxes there would be a block instrument 
for the up line, a starting signal, the necessary lever for 
working the signal, and an electrical contact treadle in 
advance of the starting signal. 

In the upper part of the block instrument is a miniature 
arm, which, when down, indicates that the line in advance 
is clear. There are also two apertures in which indications 
appear, e.g., when the word "locked " appears in the upper 
aperture in the face of the block instrument it shows that the 
starting signal is locked. When A wishes to send a train to 

B, he gives the usual bell signals, and if B is ready to 
receive the train he presses his plunger, and this causes 
the words " train-on " to appear in the lower aperture of his 
instrument, and the starting signal at A to be unlocked; at 
the same time the word " locked " in the instrument at A is 
replaced by " free. " The acceptance of the train by B further 
raises the miniature arm in the instrument at A- 

On A pulling over his starting-signal lever, the sign in his 
block instrument is again changed to " locked," and not until 
he has put his starting signal to danger and the train has 
gone over the treadle can he accept another train, nor can 
this train be accepted by B until the first train has arrived 
and gone forward, and before B can lower his starting signal 
for the train to go forward to C, the man at the latter box, 

C, must do for B what B has had to do for A- 
Proceeding now to study the working in detail, the type 

of instrument mostly in use is seen in figs. 46 to 50. 

The method of working these instruments and the indi- 
cations differ on the various railways, but the main principle 
is the same, i.e., to prevent two trains being in the same 
section at the same time. One of the best features of the 
Sykes system is its ready adaptability to suit all the exi- 
gencies of traffic working. 

The instruments are fixed on the usual block instrument 
shelf above the locking frame. There are three strong rods 
(fig. 46), S, V, X, connecting the locking instrument with 
the signal lever that has to be controlled. S is the actual 
locking rod, Y is the rod that actuates the " / rain-on " disc 
and prevents the plunger being used a second time to accept 
another train until the signal lever has been pulled and put 
back, and X is the switch rod which, by the motion of the 
lever, connects in one position (signal " on ") the line wire 
to the locking coils; and in the other (signal "off") the 
treadle wires to the locking coils, and disconnects the line 

Figs. 47 to 49 show the front and side views of the instru- 

Fig. 50 shows diagrammatically the connections between 
the instrument and the lever, so that the effect of the dif- 
ferent motions may be made clear. The electrical connec- 
tions are also shown, to better illustrate the mode of working. 
The locking instrument coils i are fixed upon soft iron continua- 
tions of a powerful permanent magnet 2, and the armature 3, 
in the locked position (as shown), is held up to the cores of 
the coils by magnetic attraction. The coils are wound, or 
the circuit joined up, in such a direction that a strong cur- 
rent from the line neutralises the magnetism induced by the 
permanent magnet, and the armature 3 (which forms a por- 
tion of an angle piece 4, with its axis at 5) pressed away 

B >C 

Fig. 45. Sykes' Lock and Block. 


Fig. 46. 


Fig. 47. Sykes' Lock-and-Block. 

Fig. 48. 

Fig. 49. 

from the cores by a strong adjustable push-off spring 6, is 
discharged, and the small wheel 7 on the opposite arm of 
the angle piece 4, sliding from underneath the catch-piece 
9 (attached to the locking rod 9), allows the rod to fall, and 
in doing so it raises the lock 10 out of the slot n, and thus 
releases the signal lever \2. 

With this diagram (fig. 50) all the operations may be 
followed. Let it be assumed that the lever and the locking 
instrument (which is really the sending portion of the 
apparatus) are at cabin A, and the plunger 13, battery spring 
14, switch hook 15, battery 16 (in a receiving instrument), 
are at cabin B, the treadle 18 being a train's length ahead of 
the starting signal worked from cabin A. 

A has a train, and gives the " is-line-clear " signal to B 
on the block bell. If the section be clear, and B has his 
starting signal at danger (he cannot do so without), the latter 
plunges, pushing off the click holding up the rod 25, causing 
it to fall, and indicating to himself " train-on " by the disc 
21 at the lower aperture (see fig. 46) in the screen, and his 
plunger 13 becomes locked. This plunge joins the battery 16 
to the line, and the rod 25 falling, severs the spring 29, 
cutting off the battery 26 working the miniature arm at A. 
The current transmitted releases the lock 10 at A as ex- 
plained, and indicates, by a disc 24 carried on the rod 9, 
" free " and raises the arm over the instrument. The signal- 
man at A then pulls over the lever (12) in the locking frame 
and lowers his starting signal, and this operation has the 
effect of raising the locking rod 9, lowering the switch rod 
20, connecting the treadle circuit to the locking coils i, and 

disconnecting the line. At the same time the lock 10 has 
again dropped into another slot 21, in the tappet 22 of the 
lever, back-locking the same in the " off " position, and indi- 
cating "locked" on the instrument. This slot 21 is so 
arranged in the tappet that, although the lever cannot be put 
right back so as to allow of another train being accepted, 
it can be put sufficiently far back to throw the signal to 
danger in case of emergency. 

On the train leaving A, " train-cntcring-section " is given 
on the bell to B, and on its passing over the treadle 18 the 
circuit is closed and the back lock is released, the signal put 
to danger and relocked, and the line wire again connected to 
the locking coils i, the treadle circuit being left discon- 

After the train has passed B and the signal there put to 
danger, the putting back of the signal lever 12 again raises 
the rods 25 and 27, carrying the " train-on " disc 21, 
changing that disc to blank, rejoining springs 29 and battery 
26, so lowering the miniature arm at A. 

Should B require to block back the line from A, he can 
do so by turning the switch hook 15 over the plunger, which, 
breaking the line (by separating the two springs 14 and 140 
and cutting off the battery 26), indicates to A " line-blocked " 
by raising the miniature arm, and at the same time locking 
his own plunger at B. 

This is an operation that should be performed when a 
signalman is going to foul a road by shunting a train across 
from, say, the up to the down line. 



Fig. 50. Sykes' Lock and Block. 

If a down train has to pass A before it reaches B, and the 
signalman at B is going to shunt an up train on to the down 
line, he ought first to turn the switch hook 15 over his 
plunger, which will at once break down the line, and not 
only indicate to A by the raising of his miniature arm that 
the down line at B is obstructed, but will lock the starting 
signal at A so that it cannot be lowered. 

The same operation should be performed before the sig- 
nalman allows a line to be fouled by a train leaving a siding 
for the main line. This, it will be remembered, was what the 
signalman at Tottenham North Junction omitted to do in 
the collision referred to in Chapter II., page 21. 

It must be noted that when the starting signal is 
lowered (fig. 50) the signal lever is locked in the " off " 
position. The advantage gained by this lies in the fact that 
when a signal lever is " over " it interlocks, by the usual 
mechanical interlocking, all the conflicting points and signals 
and, consequently, when a train has been accepted and the 
starting signal lowered, it is not possible to put the signal to 
danger again and commence shunting operations, or, at a junc- 
tion, to reverse the junction and accept a train that would 
foul the path of the first train. 

But still it might be necessary to put the signal to danger 
in order to stop the train, and so, as has already been stated, 
the slot 21 (fig. 50) is so arranged that the lever can be put 
sufficiently far back to put the signal to danger, but not 
enough to free the interlocking. This is a very important 
point, and Mr. Sykes is to be congratulated on having 
arranged to combine safety with freedom in working. 

In the illustrations already given the workings described 
have been for the usual movements at an ordinary block 

post. It is now necessary to describe the working of a junc- 

Fig. 51 illustrates an ordinary double junction, with a 
signal box at C- The adjacent box on the main line is at A 
and the one on the branch is at B. The next block-box on 
the other side is at D- 


. 51. Junction Working. 

The Sykes system, already described, is in force between 
each of the boxes named with the following important addi- 
tion, viz., that when a train has been accepted from D one of 
the home signals G, H, has to be lowered, and the train must 
have passed it before a second train can be accepted. For 
instance, should the train be going on to the branch line the 
home signal H has to be lowered, and after that has been 
done that signal cannot be put fully to danger until the train 
has passed over the treadle on the branch line. And, of 
course, if the signal H cannot be put fully to danger, it 
follows that signal G cannot be pulled off ; in fact, as long as 
the lever working the signal H is not properly home the 
points K cannot be moved. There is also a species of inter- 
locking between the block-instruments by which it is impos- 
sible, after a train has been offered to B, for another train to 
be offered to A until the first has been disposed of. 

So much for trains coming in the facing direction. 

Turning now to the other direction it will be seen that 
signal E is for coming on the main line and signal F for 
coming on the branch line. These signals are, of course, con- 
flicting the one with the other, and whilst the mechanical 
interlocking prevents both signals being " off " together there 
is nothing ordinarily to prevent the signalman from accepting 
a train from A and one from B at the same time. 

Such a mistake is not possible where " Lock-and-Block " 
is in force, as the acceptance of a train from A at once locks 
up the instrument giving communication with B and vice 

But Mr. Sykes goes very much further. 

It is one of the rules of block working that if a train 
which requires to go to A be approaching C from D, and at 
the same time a train is approaching C from B on the branch 
line, then the signalman at C must so set his junction 
points that should the train from D over-run its signal G it 
shall not come into collision with the other train. This is accom- 
plished by setting the points K so as to lie for the branch. 
In the Sykes system there is interlocking between the frame 
and the block instruments so that a train cannot be accepted 
from B unless the facing points K are set for the branch 
line, and after the train has been accepted the points are 
" back-locked " so that they cannot be moved until the train 
has been disposed of. Also, of course, if the points K lie for 
the main line then a train cannot be accepted from B- 



If a train has to come from A along the main line it can- 
not be accepted if the points J be not lying right, and after 
the train has been accepted from A it is impossible to move 
points J so as to lie for the branch until the main line train 
has been disposed of. 

Should such a junction be also a station, as seen in fig. 
52, the arrangements are somewhat modified. Here there are 
inner and outer home signals E 1 E 2 and F 1 F 2 , from A and B. 



Fig. 52. Junction Working with Station. 

The same arrangement applies, as in fig. 51, as to it not 
being possible to accept a train from B unless the points K 
are lying right for the branch, and as to a train not being 
able to come from A if the points J be not lying right for the 
main line. 

But supposing the same conditions applied throughout, 
and it were not possible to move points J after a train had 
been accepted from A until that train had passed cabin C, 
then it would not be practicable for a train from A to make 
connection with a train from B ; also it would be possible to 
let a stopping train from B, that had to come to rest in the 
station, to hold up a fast train that wanted to go from D 

to A- 

Therefore the connections between the block instruments 
and the locking frame are so arranged that before a train 
can be accepted from B, the points K must be set for the 
branch, and then, when the stopping train has arrived in the 
station, it passes over a rail contact or treadle, and after it 
has come to a stand at the inner home signal F 2 at danger, 
it is possible to put the lever working signal F 2 fully back in 
the locking frame, and then points K can be reversed. 

At a signal box where there is a connection with a 
siding, the working of the same is rendered safer by the 
use of the Sykes system. 

The mechanical interlocking prevents the siding points 
being opened when the main line signals are " off," but what 
is required is the acceptance of a train to be made impos- 
sible when such an operation is being carried out, and this, 
too, is provided by the Sykes system. 

When the lever that works the points of the siding is 
over in the locking frame, the block instrument is locked so 
that it is impossible to accept a train on the line that is 
being fouled. Similarly, when a train has been accepted, 
the lever working the points of the siding is locked, so that 
the main line cannot be fouled although the main line signal 
may be at danger. Under ordinary working, when the sig- 
nals are at danger, it is possible for the siding-points to be 
opened although a train may be approaching. 

This is made clear by referring to the diagram fig. 53, in 
which there is indicated a connection between a siding and 




^ DOyvN 


53- Siding Connection. 

the up main line, which connection is worked by lever 5. 
Should this lever be moved then it is not possible for the 
signalman to accept an up train from A- Similarly, when 
an up train has been accepted from A, then it is not possible 
for the points of the siding to be opened, although Xo. 2 
home signal may be at danger. 

This same safe working can also be applied to cross- 
over roads, but it is somewhat more complicated, because 
when a train is shunted from one line to another, one line is, 
of course, freed while the other is blocked. 

Fig. 54 illustrates a signal box B where there is a cross- 
over road worked by lever No. 6, a down home signal 2, a 
down starting signal 3, an up starting signal 9, an up home 
signal 10, and disc signals 7, 8, for going through the cross- 
over road. 

2 o 
. nntv/v I - 




Fig. 54. Crossover Road. 

The signal boxes on either side of B are A and C- 

When B plunges to receive a train from A, he locks No. 6 
lever until No. 2 home has been pulled and put back, and 
then Nos. 2 and 8 are locked until the train is disposed of by 
the use of No. 3 starting signal. 

The same method applies to the other road. When a 
train is accepted from C, No. 6 is locked until No. 10 home 
has been pulled and put back, and then Nos. 7, 10, are locked 
until the train has been disposed of by the use of No. 9 
starting signal. 

When No. 6 lever is moved to use the crossover road the 
block instruments for both lines are locked, but the use of 
No. 7 shunt back would free the down instrument, but would 
continue to lock the up instrument, and also to lock Nos. 7 
and 10 levers until the train has been disposed of by the use 
of No. 9 starting signal. The same means would apply to 
the down instrument and Nos. 2 and 8 signals, when No. 8 
disc was used to shunt a train from the up to the down line. 

Fig- 55 is interesting, as it illustrates a position where 
there is no starting signal and no shunting discs at the cross- 
over road. There is, however, a treadle on each road in 
advance of the signal box as in all the examples given. 


T- 55- Block Cabin with no Starting Signals. 
The home-signal 2 is unlocked by the acceptance of the 
train by the box C in advance, there being no starting signal. 
Similarly on the down road the home signal 4 is released by 
the acceptance of the coming train by the box A in advance. 



When the crossover road 3 is worked the pulling of the 
lever locks both plungers so that a train cannot be accepted 
from either A or C- 

Fig. 56 illustrates a case in which up and down loops 
join the main line. After a train has been accepted on the 
up main line, signal 2 is free and signal 3 becomes locked, 

Signal 'Cabin. 


Fig. 56. Junction with Loop Lines. 

and this signal remains locked in the normal position and 
signal 2 in the " over " position until starting signal 4 has 
been pulled and put back and the train has passed over the 
treadle in advance. 

In the same way when a train has been accepted along 
the up loop and signal 3 has been lowered, it is locked in the 
" over " position and signal 2 in the " on " position until 
the train has passed starting signal 4 and gone over the 

The same working applies to the down road so far as 
signals 9 and 10 are concerned. They each lock each other 
and the pulling off of the one locks the other in the normal 
and itself in the " off " position until the train has been dis- 
posed of. 

Also in this case, when a train has been accepted from A 
on the up main line, the loop outlet points 5, 5 are locked 
and so also is the crossover road 6, 6. And when a train has 
been accepted on the down main line then the down loop out- 
let points 7, 7 are locked and also the crossover road 6, 6. 

When the points 5, 5 are used to let a train out of the up 
loop, then the plunger in the block instrument by which a 
train is accepted from A is locked. Similarly, when a train is 
being admitted on to the down line from the down loop ther 
the plunger by which a down train is accepted from C on the 
down main line is locked. 

When the points 6, 6 are being used to cross a train 
from one main line to another, then both plungers are 

The working of crossover roads is rendered easier when 
there are shunting discs, as in the case illustrated by fig. 54. 

There is, of course, no need for 5, 5 to lock 6, 6, for 6, 6 
to lock 5, 5 and 7, 7, nor for 7, 7 to lock 6, 6, as all this is 
done in the mechanical interlocking. 

It is perhaps desirable to state that if these loop lines 
were passenger roads and were not provided with safety 
points then the acceptance of a train on the main line would 
lock the loop line instrument instead of locking the loop line 
points. As shown in fig. 56, and as already explained, the 
acceptance of a train on the up main line locks the points 
5, 5, so that if a train comes along the loop it cannot enter 
on the main-line and therefore it may be free of the main 
line instrument. 

The connection between the locking frame and the block 
instruments in these special cases is illustrated by fig. 57. 

Let the junction as shown in fig. 52 be taken as an 

Lever 7 

Fig. 57. Junction Working. 

example. In this case a station is at a junction, and there 
are inner and outer home signals for the up main and branch 
lines, the outer home signals being for the protection of the 
station, and the inner home signals being at the end of the 
platforms and protecting the junction. 

The up main signals are E 1 and E 3 , and the branch sig- 
nals are F 1 and F-, and the difficulty previously noticed was 
that when a train had been accepted on the up branch instru- 
ment a train could not be accepted to proceed along the down 
main line until the branch train had passed the station and 
was on its road into the next section. And this difficulty was 
accentuated when the branch train had to stop in the Nation. 

Special interlocking is therefore provided, so that when a 
stopping train has arrived in the station and the inner branch 
home signal is at danger, the facing points in the down line, 
which the mechanical interlocking has compelled the signal- 
man to set for the down branch line, could be reversed and a 
down main line train accepted. 

This object is attained as follows : 

On the block instrument shelf are two instruments, the 
one on the right being the block instrument to the signal 
box B in the rear on the branch and the one on the left 
being the lock instrument for recording the position of the 
lock on the facing points K (fig. 52 and fig. 57), worked by 
lever 7. 

Attached to the shelf carrying the block instruments is a 
small lever H working a shaft and having connections with 
both the lock and the plunger instruments. The shaft is also 
connected to the lever working the facing points by means 
of the downright rod K and the tappet M which is attached 
directly to the lever. 

When the signalman works the points (K fig. 52) by 
lever 7, the tappet M travels along with the lever and the 
notch R comes under the downright rod K, and the turning 



piece L can then enter the notch at right angles. The small 
lever H may then be pulled forward, and the turning piece L 
enters the notch R and holds tappet M. Therefore the lever 
cannot be moved from the " over " position and lever 7 is 
" back-locked." At the same time the plunger is unlocked 
by the dropping of the rod J which is also attached to the 
shaft actuated by the lever H, and the plunger S can then 
be used. 

When the small lever H is pulled the armature X is 
raised to the coil O, and the tail piece T passes under the 
angle piece P (which has been raised in the meantime by the 
movement of the rod K), and the downright rod cannot 
therefore be moved, the turning piece L is held fast in the 
tappet M, which of course holds the latter, and so, in its 
turn, the lever 7 is locked in the over position and cannot be 
put to the normal until the small lever H is put back. 

But the small lever H cannot be replaced so long as the 
armature N is held up by the coil O, and this state will con- 
tinue until a current has passed through the coil in the 
reverse direction. This current is set up by the deflection of 
a treadle G in the station on the branch line, but the current 
on its way to the coil has to pass by the lever working the 
inner home signal for the branch line, and this signal must 
be at danger. 

Given then that the inner home-signal be at danger and 
the train passed over the treadle, a current is set up which 
passes to the coil O, frees the armature N, the tail piece T 
falls, the angle piece P is free and then the small lever H 
can be put back, and the lever 7 in the locking frame can be 
restored to normal and the facing points then set for the down 
main line. 

Perhaps just a few words are necessary to explain the 
reciprocal movement, viz : how a train, having been accepted 
on the down main line, locks the up branch instrument. 

It will be noticed that the normal state of affairs is as 
illustrated in fig. 57, which is that the plunger S is locked 
by the upright rod J. This, as has been seen, cannot be 
lowered except by the small lever H, which cannot be moved 
unless the notch R is in line to receive the turning piece L, 
and that is not until lever 7 is over. As the ordinary Sykes 
interlocking provides for this lever being held after a down 
main line train has been accepted, it will be seen that when 
such a train has been accepted it is not possible for a train to 
be accepted that will cross the path of the other train. 

An arrangement something similar to that illustrated in 
fig. 57 is provided in all the other examples given. 

Spagnoletti's system. 

Throughout the whole of the Metropolitan R. and on 
portions of the Great Western R. the Spagnoletti system of 
" Lock-and-Block " is in use. It is the invention of Mr. C. 
E. Spagnoletti, who was for many years the electrical 
engineer of the latter line. 

The object attained is the same as in the Sykes system, 
viz., that the signal for entering a section shall not be 
lowered until the train has been accepted by the box in 
advance, and the signal having been lowered for a train, 
must be put to danger and the train must have gone over an 

Station A 

Station B 

Up Line 


Do~n Line 



o e 




Doirn L Ine 





o e 


H n 








Fig. 58. Spagnoletti's Instruments. 

electrical treadle at the advance box, before a second train 
can be offered. 

The block instrument is of special form as shown in fig. 
58, which illustrates companion instruments working 
between two boxes A and B- 

As seen, each instrument has two openings. Behind the 
upper one appears a screen with three signals " Lock-on," 
"Line-clear" and 'Train-on-line-going." Behind the 
lower one three signals are given " Train-arrived," 
"Line-clear-sent" and " Train-on-line-coming." The indi- 
cators are coloured as follows : 







The normal position of the upper screen is " lock-on " 
and of the lower "train-arrived." There are two keys, a 
which is white and lettered " acknoivlcdgment-of-train- 
arrived " and b (red) " train-on-line." 

A brass plunger c is provided for the purpose of taking 
off the lock of the starting signal at the box in the rear, and 
a key d at the side for giving the bell signals. In case of 
failure of the treadle to unlock the instrument a key is pro- 
vided in a glass case. On breaking the glass possession can 
be obtained of the key and this will allow access to the re- 
setting button e which will take the lock off and allow for 
the arrival signal to be given. 

In the lower left hand corner of the instrument is a 
switch / by which the instruments can be converted from 
double line working into single line working and the up 
and down block signals can then be interlocked with each 
other in case of an obstruction and single line working having 
to be put into operation. 

This is done by inserting the key kept in the glass case 
into the key hole g and turning the switch / from D to S. 
When double line working is resumed the switch must be 
again turned to D and the key taken out of g and replaced 
in the glass case. 

The mode of procedure for ordinary working is : 

A train has to travel from A to B- The signalman at A 



. Controlled by [lectricSlot 
from B . 

. Controlled by Electric Slot 
from C. 

Controlled by Electric Slot 

i , 

i fi ram passing puts N8to danger ^unlocks 
handle of other instrument! f there was one 
, 8\\ \ fo 

,-"" Tram passing puts 

' \ H7to danger an d 
7J [ unlocks handle ofb 

.'-- Tram passing puts 

_ - Wo to danger and 
1 || unlocks handle ofc 

Irain passing 
unlocks the 
^handle of d. 
i Jl 1 ' + 

1 0/5 j t^ 



, 09 
Irampassing TH 
unlocks the i 
handle of a f 

', ~ -fOU 
tlram passing puts Tfo 1 Bj 
-1 H2tajanaerana i V _ 
unlocks handle of b' \ t 

Controlled by 
Electric Slot from A :' 

' 6' 


.. y/0 | 

train passing puts TTj 
N 3 to danger and * 
unlocks handle ofc' 

Controlled by 
Electric Slot from B 

MR Train passing puts N+ TTT 1 BOg 
*P^ to danger and unlocks V 1 
^ handle of 'other mstru- D 
ment. if there itas one 
Controlled by \ 

L Worked from B Worked from A b 

-Worked from C Worked from B . c 

: C' Harked from D Worked from C l 



and releases and releases ij 
Signal N 8 Signal N 2 


and releases and releases ii 
Siqnal N 7 Sianal N 3 JP 

dj arj releases andreleases i 
F 1 Signal N 6 Signal N4 



H*- - -.0 

'--Iforij llectnc Works ffeclnc *X- 
Sloton N 2 Slot on N" 8 


'--Works Electric Works Electric- W- 
Slot on N' 3 Slot on N 7 

<* ' Works Electric Works tleclrit *& 
Slot on N 4 Slot an M' 6 


Fig. 59. Hodgson's Lock-and-BIock : Diagram of Lines. 

gives the prescribed bell-code signals to B on key d. If the 
latter be ready to accept the train he presses in his lock 
plunger c, which action locks up his plunger so that he can- 
not use it to accept a second train. It also records on the 
lower screen of the instrument at B " line-clear-sent " and 
at A it shows " line-clear " in the upper opening. This action 
also frees the starting signal at A and allows the signalman 
there to lower that signal for the train to go forward to B- 

On the train leaving A, the signalman there must first put 
his starting signal to danger and give the " train-entering- 
section " block signal on key b, and this causes the upper 
screen of the instrument at A, which went to " lock-on " 
when the starting signal was put back, to indicate " train- 

The instrument at B will, at the same time, show " train- 
on-line-comlng " in the lower opening. When the train ar- 
rives at B the signalman there gives " train-out-of-section " 
in the usual way. A will then plunge his " acknowledgment- 
oj-train-arrived " key a, when " lock-on " will again appear 
in the upper opening of his instrument, and it is not until the 
train has passed B and gone over an electrical treadle there 
(which causes " train-arrived " to be shown on the block 
instrument at B) that the signalman there is in a position to 
accept a second train. 

It is claimed for this system that it can be applied to any 
form of block instrument : that it protects ail movements out 
of loops and sidings, and from one main line to another 
when crossover roads are used : its applicability to junctions : 
that the signalman at B, when he sees the signal " train-on- 
line-coming," knows that the signalman at A has pulled off 
his starting signal, and if that signal is not afterwards put 
to danger, a second train cannot be accepted by B. 
Hodgson's System. 

In this system the instruments are of the ordinary form, 
as is the plunger, but the commutator by which the electric 
signals are exchanged is of special construction. 

There is the usual interlocking between the block instru- 
ments and the locking frame, also an electrical treadle con- 
tact maker. In addition to these usual fittings, each signal 
controlled has an electrical slot. 

Fig. 59 is a diagram of a line controlled bv four signal 
boxes A, B, C, D. As the starting signals at these boxes are 
the only signals affected, they are the only ones indicated 
and are numbered i 8. 

The block instruments are a, b, b 1 , c, c 1 , d, the instru- 
ments working between A and B being a and b, between B 
and C being b 1 and c, whilst c 1 and d work between C and D. 

The electrical treadles in advance of each box are indi- 
cated at 9, 10, ii and 12, for the down line, and 13, 14, 15 
and 16, for the up line. 

It will 1 > seen from the sketch of the block instruments 
that the electrical indications take the same shape as a two- 
arm outdoor signal, the upper arm being for the up line, 
and the lower arm for the down line. So long as the minia- 
ture arm is raised the line is " blocked," and when the arm is 
lowered the line is "clear." 

The plunger is in the lower part of the instrument on the 
left-hand, and the aperture on the right is where the screens 
" line-clear " and " train-entering-section " appear. 

Connected with the plunger is a down-right rod leading 
to the locking frame. This is to lock those point levers 
which have to be interlocked when a train has been accepted. 
This rod is not shown in fig. 59, but it is in the larger draw- 
ing fig. 66 of which the description is given later. 

Each signal controlled is provided with an electrical slot 
which is connected on to the line-wire actuating the block 

The working is as follows : 

If an up train has to go from B to C, the signalman at 
B gives the recognised bell signals on the plunger of instru- 
ment b 1 . If C be prepared to accept the train, he gives the 
prescribed bell-signal on the plunger of his instrument c, and 
then turns the handle of his instrument from left to right. 
This operation has two results. In the first place, the down- 
right rod locks up conflicting point levers so that no shunt- 
ing can be performed after a train has been accepted, and if 
C were a junction, the junction would have to be properly 
set before a train could be accepted, and after its acceptance 
the road could not be altered. It also follows that if any 
point levers be over for shunting operations, the down-right 
rod would be back locked and could not be moved nor the 
handle turned to accept the train. 

Then, secondly, the turning of the handle from left to 
right puts the instrument c into such a condition that on the 
signalman at C plunging to B, " line-clear " is given, the 
upper arm of b 1 falls, and the electric slot on signal 7 (the 
starting signal for the up line at box B) is taken off, so 





Fig. 60. 

Fig. bi. 
Figs. 60-62. Hodgson's Block Instrument. 

Fig. 62. 

allowing the signalman there to pull the signal " off " in the 
ordinary way by means of a lever in his locking frame. 

After the train has gone into the section B C, the signal- 
man at B puts signal 7 to danger again, and it cannot be 
lowered a second time. He also sends " train-entering-sec- 
tion " signal to C, when the man at the latter box turns his 
handle from right to left. It will not, however, go all the 
way and consequently the point levers continue locked. 1 he 
turning of the handle puts the instrument c into such a con- 
dition that on the signalman at C again plunging, the upper 
arm of instrument b l is again raised, the screen shows 
" train-entering-section," and the electric slot on signal 7 
goes on. The handle of c becomes locked in that position 
and it cannot be put completely to the left to release the 
point levers, nor to the right to give " line-dear," until the 
train has arrived and gone into section C D, when it will 
pass over electrical treadle 14, which then frees the handle. 

The signalman at B might omit to put his starting signal 
7 to danger. This contingency must be provided against, 


and it is done by means of the electrical treadle 15, which 
puts the signal to danger, and it cannot again be lowered 
until the whole operation for the train has been gone through 
on the block instruments. The signalman at C must turn 
his handle from left to right, and then again from right to 
left, and, as has been seen, this cannot be fully done until 
the train has gone over treadle 14. It is clear that a second 
train cannot enter a section until the first one has been dis- 
posed of, nor can a train enter a section until it has been 
accepted by the box in advance. 

The details of the Hodgson block instrument are illus- 
trated by figs. 60 to 62. Fig. 60 is an end view, fig. 61 an 
inside front view, and fig. 62 a front view of the upper por- 
tion, and an inside view of the back of the lower portion. 

The plunger is a, the handle b, the coils c c are for the 
indicator arms d d, the coils e c are for unlocking the handle 
b. The screen / shows what electrical signals have been 
received from the box in advance, and cause " line-clear " 

Fig. 63. Normal " Line Blocked." Fig. 64. Reversed " Line Clear." Fig. 65. Train in Section " Line Blocked." 

Handle of Hodgson's Block Instrument. 



and " train-entering-s action " to appear before the aperture 

\\ hen the handle is in the normal position, the pressing 
of the plunger a joins springs h and /, and fe and I, and 
severs the contact between k and m. The effect of this is to 
ring the bell at the box in the rear. On the handle being 
put over to the right, springs h and I and fe and j are joined 
up, springs fe and m still being separated. The consequence 
is that when the plunger is used, a current is sent that indi- 
cates " line-clear " in the manner already described. 

On the handle being placed in the middle position, the 
original contact is made, and the effect of this is to show 
" train-entering-section" after the plunger has been pressed. 

Figs. 63-65 illustrate the three positions the handle can 
be placed in : Fig. 63 shows the normal position when the 
handle is to the left, and the upper arm of the block instru- 
ment in the rear box is up. 

Fig. 64 shows the position when the signalman turns the 
handle from left to right in order to accept a train from the 
rear, so lowering the upper arm of the block instrument and 
taking off the electrical slot. 

Fig. ($ illustrates the position when the train has been 
signalled as passing the rear box, and the signalman has 
raised the upper arm and put the electric slot on. This 
position of the handle has already been referred to. It will 
be noticed that it has not gone fully over, and there it 
remains until the train has gone over the treadle in advance 
of the box which then allows the handle to be moved com- 
pletely to the left. 

On the shaft working with the handle is a quadrant plate 
a, having three indentations b 1 , b 2 , fc 3 , fitting the spring c 
which holds the plate in position. 

There is also a three-arm lever d, one arm e of which 
limits the movements of the lever between the stops / f. 

\Yorking on a separate centre is the hook g attached to 
the crank h, and on the lever d being moved, the arm / 
comes into contact with the arm fe of the crank h, and raises 
the hook g to the position shown in fig. 64 ; on the hook g 
being raised, the stud I forces its way past the electro- 
magnet m and comes to rest on the top of it and is held 

Fixed on still another centre is the tail piece n, on which 
are fixed the screens (/, fig. 61), and this is kept up by an 
arm of the crank h, and so in passing from the position 
shown in fig. 63 to that shown in fig. 64, the tail piece n is 
freed and falls. 

Y\ hen the handle is moved to the middle position to show 
" train-entering-section," the state of the instrument is as 
seen in fig. 65. There is very little difference between the 
two except that it will be noticed that the arm o of the lever 
d has passed the hook g on the tail piece n, and consequently 
the handle b cannot be moved to the right to again show 
" line-clear," and the hook g being kept up by the stud I, 
the arm / is held and consequently the handle cannot be 
moved further to the left to give "line-clear." This cannot 
happen until the train has gone over the treadle at the box 
in advance, when the electro-magnet m is cut out and frees 
the stud /, allowing the hook g to fall, and the crank h 

Fig. 66. 

End View showing Connection between Hodgson's Block Instrument 
and Locking Frame. 

comes out of contact with the arm / of the lever d, and 
consequently the handle can be put fully to the left, and 
" train-out-of -section " signal sent. 

On the hook g falling, the projection p enters the aper- 
ture r of the tail piece n, and raises it and causes the screen 
to show normal. 

Fig. 66 shows an end view of the connection between 
the block instrument and the locking frame. 

The pattern here shown is in connection with the Saxby 
& Farmer " rocker " locking-frame, but it can be applied 
equally readily to their present frame or any other form of 




Fig. 67. Normal Condition. Fig. 68. 

Signal Ann " Off." Fig. 69. 

Hi d ;snn's Electric Slot. 

Signal Arm to Danger independently of Signalman. 

It will have been noticed that there is no connection with 
the locking-frame to unlock the starting-signal, as there is 
in most other systems. 

In the Hodgson system this end is attained by the elec- 
tric slot on the starting signal. 

The connection (fig. 66) is simply to lock up the frame 
when a train has been accepted, also to prevent the handle 
of the block-instrument being moved to accept a train if 
any conflicting lever be over. It also guarantees that at a 
junction the points arc properly set, and before a train can 
be accepted from the box in the rear on the branch, the 
trailing points must be set for the branch, and the act of 
pulling over the lever to set the road would at once lock the 
main line instrument for the converging road. 

Figs. 67-69 illustrate the electrical slot ; fig. 67 shows 
the normal condition and the arm ready to he lowered; fig. 68 
shows the arm " off," and fig. 69 the arm after it has 
gone automatically to danger by the passage of the train 
over the electrical treadle, or by the sending of the " train- 
entering-!, eclion " signal, and independently of the signal- 
man's action. 

The slot is attached to the signal post by means of the 
plate a. There is the usual balance lever b with the signal 
wire c, and the upright-rod d. The rod d is kept in line by 
two links e 1 e~, one end of each of which is attached to the 
rod, and the other end to the plate on the signal-post. 

Also on the upright-rod is the toggle /, one side f l of 
which is coupled to the balance lever b and the other side f 
vo the lever g by the link g l . 

On the plate a is the box h carrying an electro-magnet 
and on the lever g is an armature /. 

The normal condition of the slot is as illustrated in fig. 
67, and when the signalman at the box in advance accepts a 
train a current passes through the coils of the electro-mag- 
net h and attracts the armature ; and so forms a fulcrum for 
the toggle /, and on the balance-lever b being raised by the 
signal-wire c being pulled the signal is pushed " off " as 
seen in fig. 68. 

It has already been stated that the signal will go auto- 
matically to danger, in case the signalman omits to put the 
lever back, by 

A. The signalman sending the " train-entering-scc- 
tion " signal on the block instrument. 

B. The train going over the electrical treadle at the box 
in advance, in case the signalman at the rear box having 
not only omitted to put the signal to danger but omitted to 
send the " train-entering-section " signal. 

In either of these events an opposite current passes 
through the coils h and the armature ; is no longer attracted. 
The lever g cannot sustain the weight imposed upon it and 
therefore it falls and the upright-rod d comes down too, and 
the signal goes to danger as seen in fig. 69. 

Before the signal can be again lowered the armature ; 
must be put in contact with the coils h and this is done when 
the balance-lever b goes to normal as it raises the short rod 
k, actuating the tail piece I which comes into contact with 
an arm of the lever g, and raises it from the position illus- 
trated by fig. 69 to its normal position as in fig. 67. 

The special treadle used in connection with the Hodgson 
system is described in Chap. V. (fig. 1 13). 



Mr. Hodgson claims many advantages for his system, 
but the special ones demanding attention are 

1. One wire works both audible and visual electrical 
signals for both up and down lines ; the electric current is 
transmitted by simple means ; and there being only one 
handle carrying the spring pjunger, there is no liability of 

2. Signal invariably put up to danger behind trains, as 
two men can put it up, and if they fail to do so the train 
itself does it, and then the signal cannot be lowered again 
until again released by the man at the signal-box in advance. 

3. Economy of battery power, the connection to battery 
being cut off directly the work is done, and no battery in use 
to the slot till the spring catch of the lever is grasped for the 
purpose of working the signal, so ensuring the batteries 
lasting the maximum time. 

4. Any one part of the system can be omitted without 
in any way detracting from the advantages of what remains. 
For instance, the use of the electrical slot alone, or the inter- 
locking between the handles of the block instrument and the 
locking frame, or the electrical treadles. 

This is an important characteristic of these appliances, 
that each can be used separately and independently of' the 
other, each attaining fully its own particular object; or all 
can be used in combination, forming a perfect whole; the 
separation of one appliance from the others only does away 
with what is attained by the particular part removed, leaving 
all the other parts working as before. 

In some systems, if a part be removed, the harmony of 
the whole is so affected that the efficiency of what is left is 
impaired. This result is here avoided, and a system is pro- 
vided which, if the whole is not adopted at once, can easily 
be expanded hereafter by the simple addition of other parts, 
and without altering what is already in use. 

Langdon's System. 

This was invented by the late Mr. \Y. Langdon, formerly 
the telegraph superintendent of the Midland R., and has 
been in use for several years at various places on that com- 
pany's system. 

In designing the apparatus the inventor had in view the 
desirability of retaining in use the existing form of block 
instrument, and also of not disturbing the method of block- 
signalling the signalmen were accustomed to. In this, Mr. 
Langdon was successful. 

The instrument is illustrated by fig. 70.* It is of the 
same form as the ordinary block-instrument in use at the 
present time, except in one particular, which is, that the 
handle b has small levers a a 1 to keep it in position when 
turned to the right or left, instead of its being " pegged " 
there as used to be the Midland R. Company's custom. The 
lever a is used when the handle b is turned to the right 
(position c) to give " line-clear," and the lever o 1 when the 
handle is turned to the left (position d) to give " train- 
entering-section " signal. 

The bell signals are given on a separate gong instrument, 

* Figs. ~o, 71, and 72 are reproduced, by special permission of the 
author, from The Application of Electricity to Railway Working, by \V. E. 
Langdon. London : E. & K. N. Spon, Limited. 

Fig. 70. Langdon's Block Instrument. 

and the needle for indicating the state of the line is deflected 
in the usual way. 

Fig. 71. Lock on Block 
Instrument Handle. 

The lever a 1 , in addi- 
tion to holding the 
handle b when turned to 
position d, also connects 
up extra battery power, 
which actuates a lock on 
the lever working the 
starting signal at the 
signal box in the rear. 
The lever a holds the 
handle in position r, and 
is connected to the crank 
e as shown in fig. 71, 
which, in turn, moves 
the quadrant /, which is 

Fig. 72. Lock on Starting Signal. 

kept in position by the spring g. The movement will be 
better understood by referring to both figs. 70 and 71, the 
latter being an enlarged view of the lock on the instrument 
handle b. 

When the lever a is pressed down, the mechanism as- 
sumes the positions shown by dotted lines, and on the quad- 
rant / moving to the left it lifts the lock /z which in turn (fig. 
70) raises the hook i and causes it to rest on the top of the 
spring fe, forming an armature to the coils /. 



Fig- 73- I.angdon's Lock and Diagram of Connections. 

Working on the spindle b 1 connected with the handle b 
is the bar b 2 , and on the lock h being raised the lock passes 
to the front of b 2 , and so the latter is " back-locked " and 
consequently cannot be moved, nor of course can the handle 
b. Meanwhile the lock h is kept up by the hook i being 
engaged with the armature k, and these can only be freed 
by a current passing through the coils /. The necessary con- 
nection for the electrical current is made by an electrical 
contact, fixed in advance of the starting signal, being de- 
pressed, and thus the armature k is attracted, the hook i 
falls, the lock h becomes free of the bar b 2 , and so the handle 
b can again be used. 

Fig. 72 illustrates the lock on the starting signal lever at 
the signal box in the rear. 

A sliding bar a is attached to the lever b in the locking 

frame, and passes up and down in the 
casting' c. Secured to the casting c is the 
electro-magnet d, to the armature e of 
which is attached the lever /, having a 
cross-bar g which enters into the notch h, 
cut in the slide bar a and the sides of the 
casting c; consequently, so long as the 
armature e is not attracted, the lever 
working the starting signal cannot be 

It has already been stated that when 
the lever o l (figs. 70-71) is depressed it 
connects up additional battery power, 
and this, passing through the electro- 
magnet d, attracts the armature e, so 
lifting the cross-bar g out of the notch h 
and allowing the starting signal to be 

It is therefore apparent that unless the signalman at the 
advance box accepts a train the starting signal cannot be 

locking Cods 
Starling Signed 

73 shows the electrical connections between two 
boxes B C. On the left of the instrument at C is the lock on 
the starting signal illustrated by fig. 72. 

The starting signal is supposed to be " off " and a train 
passed, and it will be observed that should the signalman at 
C attempt to give the " line-clear " signal to B for a train, 
the current would be short-circuited and the signal could not 
be given. On the signalman reversing his lever the lock 
would enter the down-rod as seen at X and the short circuit 
would be broken and " line-clear " could be given for a 
second train. 

Fig. 74. Langdon's System. Diagram of Connections for Switching out a Box. 



Fig. 74 is a diagram of the electrical connections where a 
signal-box can be switched out. A and B are the line-wire 
terminals and V and Z the battery terminals. When the 
lever (cr, fig. 70) is turned, the switch (d l , fig. 74) is also 
turned, and thereby additional battery power is connected 
through the terminal \V. The position of the corresponding 
lever (a, fig. 70) is shown at d (fig. 74). When the switch 
shown in the centre of fig. 74 is turned the open contacts are 
closed and the closed ones are opened. 

Evans' System. 

It may have been noticed that in all the systems as yet 
explained, no train can be accepted at a signal box if any- 
other train is standing on, or fouling, the line on which the 
second train has to travel. 

The consequence is that absolute block working is 

Some railway companies do not, however, adopt absolute 
block working for their goods trains, and when a goods train 
is shunting at a signal box they allow a second goods train 
to approach the home signal, the distant signal being kept 
in the " on " position. 

Similarly, when one goods train is standing at a starting 
signal or advance starting signal awaiting " line-clear " from 
the signal box in advance, a second goods train is allowed 
to be accepted up to the home signal. This cannot be done 
under a " Lock-and-Block " system applied to absolute block 

Xor at a junction can two goods trains be allowed to 
approach a converging point. 

But were such facilities allowed the working of the line 
would be more expeditious, and as the exceptions do not 
apply when either or both of the trains are passenger trains, 
the risk run is reduced to a minimum. 

Some railway companies consider that it would be im- 
possible to work their goods and mineral traffic if all such 
trains had to comply with the requirements of the absolute 
block working, and it will readily be understood that a great 
saving in time is effected if, when a goods train is shunting at 
C, a second train can be accepted from B, and then whilst a 
third train stands at the starting-signal at B, a fourth can be 
accepted from A. 

On most lines these exceptions are provided for under 
clause 5 of the standard block regulations : " Scction-clear- 
but-station-or-junction-blocked." When this rule was first 
introduced, its use led to many serious accidents, and owing 
to the strong comments of the Board of Trade Inspectors, 
its application has now become very limited, and 99 per 
cent, of the instances in which it is now used are for goods 
trains under the circumstances already mentioned. 

It will have been understood from the explanations and 
descriptions given of the various systems of " Lock-and- 
Block," that when a train is shunting, a second train cannot 
be accepted, nor unless the first train has gone into a section 
in advance, or is otherwise completely disposed of, can a 
second train be accepted. This is regarded as fatal to 
" Lock-and-Block " by those companies who require exemp- 
tions from "Absolute Block " for their goods and mineral 
trains, and who provide for such exemptions by either the 

Fi &- 75- All-right and Caution Signal. Fig. 76. Block Instrument. 
Evans' Lock and Block. Evans' System. 

" section-clear-but-station-or-junction-blocked " rule, or by 
modifications of the " line-clear " rule. 

There are several objections that can be raised against 
the section-clear-but-siafion-or-junction-btocked rule. 

For instance, the verbal communications may be misun- 
derstood, and the driver might deny he had ever received 
any warning. Then the rule requiring a driver that has 
passed the home signal and is standing at the starting signal 
or the advance starting signal to proceed cautiously whether 
the road be clear or not, must necessarily lead to some time 
being lost in running. Further, the signalman at the cabin 
in advance has no guarantee that the signal has been acted 
up to, and lastly, the driver gets the same form of semaphore 
signal for " caution " as for a " clear " road. 

The Great Central R. and the Wirral R. have met these 
difficulties by the adoption of the Evans' system of " Lock- 
and-Block," which, while giving all the security of an 
absolute " Lock-and-Block " system, can be modified into 
a " Permissive " system, whereby a second train can be 
admitted into a section, provided the first train has arrived 
and has passed within the security of the home signal. 

This system has the further advantage that it plainly 
shows to the two signalmen concerned what signa, " Abso- 
lute " or " Permissive "has been sent and what signal has 
been received. 

There can also be no dispute between the signalman and 
driver as to the "caution" signal being exhibited, as the 
fixed signal for entering a section under "caution " is dif- 
ferent to that given by the ordinary starting signal. 

F 'g- 75 illustrates the form of the " caution " signal as 
distinguished from the ordinary " all-right " signal. The 
upper arm is the starting signal and the lower arm is the 
"caution " signal, a difference that is readily noticed both 
by day and night; by day by its shape, and at night by the 
lamp being on the inside of the post, and also that when 
at danger the light is obscured and a green light comes into 
view when the arm is lowered. 



Both these arms (the "starting-" and "caution") are 
worked by the same lever, which cannot (in common with other 
systems) be moved without permission from the box in ad- 
vance, the granting of which permission determines whether 
the " starting- " or "caution" arm is to be lowered. This 
is attained by the electric current actuating a selector (under 
the signal-box), which engages in one of two blades, one 
blade being attached to the starting signal and the other to 
the " caution " signal. When the permissive authority is 
sent, the selector is drawn into the starting signal blade and 
holds it so that when the lever is pulled, only the " caution " 
blade will move and that signal is lowered. When the line 
is clear through to the next section but one, say from 
A to C, the signalman at the box in advance, B, will send 
the selector into the opposite blade and allow the top arm 
to be pulled off. 

The block instrument in B is practically of ordinary con- 
struction, except that there is provided (over the instrument 
handle) a switch, which determines whether a train is to be 
accepted under "Absolute " or " Permissive " working, and 
in the face of the instrument above the dial is an aperture 
behind which is a screen upon which are the words 
"Absolute" and "Permissive." These work with the 
switch and show to the signalman what he has done. 

Fig. 76 shows the instrument in signal-box B which con- 
trols the entrance of trains into the section A-B. 

The handle a and the dial b are of the usual form ; the 
only departures being the switch c and the indicator d. 

On a train being offered to B and his being able to accept 
the same with a clear road he first turns the switch c to the 
left for absolute working and indicates "Absolute " on the 
screen d. He then turns the block instrument handle a to 
the left, indicating " line clear," and pegs it there. 

This action unlocks the starting signal lever at A and the 
switch c having been put in the "Absolute " position the 
current flows through one set of coils and places the selector 
on the starting-signal lever, so that on the signalman at A 
pulling over that lever the upper arm is lowered. 
When this lever is pulled the electric lock is broken down 
and so prevents " line-clear " being given for a second train 
until the first has been disposed of. This is done by the 
signalman at A giving on his block bell the " train-entering- 
section " signal when the signalman at B turns his handle 
to the right (fig. 76) showing " train-entering-section " on 
his own dial and on that of the instrument at A. This action 
again connects up the unlocking coils on the starting signal 
at A, so that when the signalman at A puts back his starting 
signal lever the normal condition is resumed so far as that 
box is concerned. 

At signal box B there is a connection between the A-B 
block instrument and the locking frame, and the home sig- 
nal lever in particular. 

When the signalman there pulls over his home signal 
lever, a connection is made in the locking frame whereby 
when the home signal is put back an upright-rod from the 
locking frame to the block instrument is raised, and comes 
into contact with a lever working on the same shaft as the 
switch c, and which turns the switch to a midway position. 

W T hen the signalman at B turns his instrument handle to 
the " train-entering-section " position, the handle auto- 
matically locks itself inside the instrument, and is so kept 
locked until the home signal lever is put back. 

The switch c remains in its midway position, and cannot 
again be turned to "Absolute " until the starting signal 
lever at B has been pulled off for the train to enter B-C 
section, and on that signal being put back to danger when 
the train has passed, the upright-rod falls and allows the 
switch c to be moved. 

Should the train not have gone forward into the section 
B-C, but be still standing at the starting signal at B, and 
A wants to send forward another train, B will find he cannot 
turn his switch to "Absolute," but that he can turn it to 
" Permissive." On doing this, he sets the selector on the 
starting-signal lever at A into the other position, so that, 
having received " line-clear," the signalman there may pull 
over his lever, but the "caution " signal will this time be 

The breaking down of the electric lock is done first, as 
in the "Absolute " system, and by the use of special inter- 
locking in the frame at B, the starting signal there must be 
lowered for the first train and put back before the home 
signal can be lowered for the second, and not only must 
that be done, but the latter signal must be put to danger 
again before a third train can enter the section. 

A special arrangement is provided at A, whereby when 
the starting signal lever is pulled off for the " Permissive " 
position, the distant signal lever is not freed but kept at 


Blakey and O'Donnell's system. 

The Great Northern R. are trying this system on a few 
sections. It not only controls and interlocks the starting 
signal by means of the block-instrument actuated from the 
next box, but it automatically records the " train-entering- 
section " signal on the block-instruments in both boxes 
when a train enters the section in case the signalman has 
omitted to do his duty in that respect. 

It provides also for a situation that may frequently arise, 
viz., should a signalman accept a train from the box in the 
rear, and wish, for some reason, to stop its coming, he can 
turn a switch in his block instrument and the starting signal 
at the box in the rear is thrown to danger. 

Fig 77 shows the arrangement. The block instruments 
are of the usual pattern. There are four electrical contact 
treadles, a l a? being at A and b l b~ being at B. 

The starting-signal c at A is provided with an electrical 
slot or reverser, so geared that the arm cannot be pulled 
to the all-right position, even when the signalman at A 
actuates his starting signal lever d, unless the line be clear 
to B and the train has been accepted by the signalman in 
that box. 

As seen in fig. 77 the section from A to B is clear, and so 
when A offers B a train, the latter accepts it and turns his 
needle to "line-clear," registering a similar signal on the 
instrument at A- 

On the signalman at the latter place grasping the clasp 
handle of the lever d and raising the catch-block e, the lower 



Fig. 77. Blakey & O'Donnell's Lock-and-Block System. 

spring- f 1 comes into contact with f and places the electrical 
slot on the starting signal c in gear and the signal can be 
lowered. When the train passes over the electrical contact 
treadle a 1 and depresses the same, the contacts on the slot 
are reversed, and the signal goes automatically to danger, 
and the current so broken down is not again set up until the 
train has gone over treadle b- and " line-clear " is given for 
a second train. 

In the event of the signalman at K omitting to send the 
" train-entering-section " signal, this is done automatically 
through the depression of electrical treadle a 2 , when a cur- 
rent, passing through the coil g, attracts the armature h 
and joins up the contacts t and brings the battery k into 
operation, whereby a strong current passes through the 
instruments and the needles are deflected to " train-entering- 
section," and there they remain until the train passes over 

Fig. 78. Sykes, Jr., & O'Donnell's Apparatus. 

treadle b 1 , when the contact is broken and the needles leave 
the " train-entering-section " position. 

The electrical slot on signal c is so constructed that it is 
only in gear when the instruments show " line-clear," and 
it will therefore be understood that should the signalman 
at B want to stop a train coming from A for which he has 
given permission, all he has to do is to turn the needle from 
the " line-clear " to the midway or normal position and the 
signal c will at once go to danger regardless of whether the 
lever d is over or not. 

The electrical contacts a'o 2 and b l b- may be contained in 
one or in separate treadles. 

Messrs. Sykes, Junr., and O'Donnell invented an ar- 
rangement, illustrated by fig. 78, which seems to be a modi- 
fication of the apparatus just described. 

The object of this invention is to cause the needles of the 
block instrument to go to the "line- 
blocked," or central position, when the 
train goes over the out-going treadle, in 
those cases where the signalman has 
omitted to send the "train-entering- 
section " signal, and the train has auto- 
matically deflected the needles to that 
position; and it is attained as follows : 
When the train depresses the electrical 
treadle a the relays b c are energised, 
and the armature of b is forced from con- 
tact d to contact e so breaking down the 
current passing through the instruments 
at A and B, and allowing the needles to 
resume their vertical position and indicate 

The armature of c is forced from con- 
tact / to contact g, which breaks down all 
communication on the instruments from 
B to A until the signalman at the former 
place has turned the instrument handle to 
the central position, which brings spring 
h into connection with contact i (see 



D * 9 

A B 

Fig. 79. Ferreira & Pryce's Lock-and-Block System 
diagram), setting up a current from B 2 , through relay c, and 
so forcing the armature from contact g to contact /, and so the 
operation resumes its normal position and communication is 

Ferreira and Pryce's System. 

Mr. Ferreira (in association with Mr. Pryce, locomotive, 
signal and telegraph superintendent of the North London 
R.) designed a system of " Lock-and-Block " which to be 
properly understood requires to be studied in connection 
with the Pryce-Ferreira block-instrument (figs. 21-22). 

A diagram of the connections between three signal 
boxes, A, B, C, is given in fig. 79. 

In the centre of the instrument there is a commutator a, 
which is moved by a circular handle in the front of the 
instrument. To the lever b in the signal box is attached a 
tappet c which is normally locked and is unlocked when a 
current passes through the coils d d, which are above the 
tappet, thereby allowing the signal to be pulled off. A rod 
actuating a circuit changer is attached to the lever. The 
attachment is seen better in fig. 80 and the circuit changer 
is shown at e in the bottom left hand of fig. 79. When the 
lever is fully over and the tappet has travelled its full length, 
a contact is formed inside the circuit changer whereby a 
current is set up and the lock / again falls into the tappet, 
" back locks " the signal in the " off " position, and there it 
remains until another current is set up by the deflection of a 
rail contact g which raises the lock out of the tappet and 
allows the signal to be put to danger. 

The apparatus is so arranged that before a signalman at, 
say, B can give " train-out-of-section " signal to A, the train 
must have gone over an electrical treadle at his box and 
into the B C section. 

At junctions the lock on the lever is reversed and the 

Fig. So. Mechanism of Circuit Changer : 
Ferreira & Pryce's System. 

signal is normally free. The levers of those signals, which 
should be locked when a certain train is accepted, are locked 
when the instrument shows "line-clear." 

This form of instrument also shows an additional signal 
on the screen to the usual "line-closed," "line-clear," and 
" train-entering-section," and which is " train-passed- 
treadle." It indicates to the signalman that his starting sig- 
nal lever is free and that he can now put it to danger. 

The Ferreira and Pryce system is in use on the East 
Indian Railway. 

Wood's System. 

Mr. Wood, managing director, Messrs. McKenzie & 
Holland, Ltd., has designed a very simple arrangement 
which goes a long way towards a complete installation of 
" Lock-and-Block,'' inasmuch as by it a signalman is pre- 
vented from using his block-instrument plunger until certain 
movements in the locking frame have been carried out. 

The arrangement is illustrated by fig. 81. Passing 
through the block instrument is an upright-rod a, which has 
secured to it a contact piece b, against which bears the 
spring c, which is attached to, but insulated from, a bracket 
d in electrical connection with the line wire e. Immediately 
to the right of contact piece b (as seen when looking at the 
block instrument from the front), is a spring contact piece / 
secured to the upper end of a lever g, working on the stud h 
attached to the instrument case, and to which is secured the 
battery wire *'. 

To the lower end of the lever g is attached a link k, the 
other end^>f which is connected to the plunger bar I of the 
block instrument, and the plunger i. When the signalman 
plunges the instrument, the lever g is moved and the con- 
tact / is brought under, but does not touch the contact b. 



Fig. 81. Wood's Lock-and-Block System. 

Carried on a support n is a catch o, which is connected to 
the upright-rod by a pin at p, and this is kept in position by 
the stud q on one side and the spring r on the other. After 
the plunger has travelled a certain distance it comes into 
contact with the catch o, and forces it off its support n, and 
consequently the upright-rod a falls, the two contacts b and 
/ are brought together and the circuit is completed, and the 
electrical signal, for which the signalman has plunged, is 
sent forward by line wire e. 

The plunger returns to its normal position immediately 
by spring power in the usual way, and carried with the 

plunger is the lever g, and consequently the two contacts b 
and / are separated and the circuit closed. On the plunger 
returning to its normal position it is locked by the upright- 
rod a, and so the signalman cannot again plunge until the 
upright-rod has been raised to its normal position by the 
actuation of a certain lever in the locking frame. 
Siemens Brothers' System. 

In fig. 82 is shown an arrangement designed by Siemens 
Bros. & Co., Ltd., and Mr. Ferreira. 

This has for its object the prevention of a starting signal 
lever being pulled over and the signal lowered until permis- 
sion has been given by the signal box in advance on the 
block instrument for the train to approach. Or the arrange- 
ment can be applied to the prevention of the signal being 
lowered until the train has gone a prescribed distance and 
passed over an electrical contact. 

One of the neatest features of the arrangement is the 
scheme whereby the signal lever must be pulled fully over 
before it can be reversed, and put fully back before it can be 
pulled again. 

Attached to the catch rod P is a bell crank I 1 working a 
tappet a, which has three shallow Botches a'a 1 * 1 and a deep 
notch a 2 on the upper side and two shallow notches a 3 a 3 and 
two deep notches o 4 a 4 on the lower side. 

A rod b passes down from the instrument shelf to the 
lever which is held down by the armature of e resting on 
the lug b 2 . 

A pawl c engaged in the notch o 2 holds the tappet a, but 
when the armature is attracted owing to the magnet e being 
energised through the signalman at the advance cabin 
accepting the approaching train on his block instrument, 
it frees itself from the lug 6 2 and the rod b rises, being 
forced up by spring x. A pivotted lever b 1 is connected to 
rod b and the head of this comes in contact with lug c 1 on 
pawl c and so lifts the pawl clear of notch a 2 and allows 
the signal lever to be pulled over. 

Fig. 82. Siemens Brothers' Lock-and-Block System 



On the pawl c rising the tail of pawl d is freed and, 
urged by its spring, the head comes under stop b 3 and so the 
signalman cannot press lever b back to normal. 

Under the tappet a is a catch / which enters the notches 
a 3 a 3 as the lever is pulled over, and it is this that prevents 
the lever being prematurely replaced without being pulled 
fully over, but when the lever is back properly the catch / is 
in the notch a 4 , which allows it to be reversed as the 
lever is replaced to normal. When reversed the catch / will 
engage against the pivotted lever b 1 directly the backward 
movement of the lever is commenced and the top of b 1 is 
forced from under lug c 1 and the pawl c is allowed to fall. 
As the tappet travels the pawl c drops into the notches 
a^a 1 , but as they are inclined on one side the pawl c rides 
out again, but if the movement of the signal lever were 
stopped and the signalman attempted to get the lever a 
second time without permission from the advance box, the 
pawl would engage against the square ends of notches a l a l a l 
and hold the lever. 

When the signal lever is fully back the pawl c enters the 
deep notch a 2 , and pressing on the tail of pawl d takes the 
latter from under stop b 3 and allows the signalman to press 
down lever b, when lever i 1 will again get under lug c 1 
and the armature rest on lug b 2 until the magnet e is 
again energised by a second train being accepted. 

McKenzie and Holland's System. 

Having been agents for the Sykes system for many 
years, McKenzie & Holland have had considerable expe- 
rience with " Lock-and-Block. " 

Figs- 83 and 84 are diagrams of the circuits in two 
signal boxes, A and B, and of the locks and indicators. 

The signal lever tappet a has in it a notch a 2 into which a 
lock b falls. The lock is lifted out by one end of the lever 
b 2 , which is centred at b 3 and has its other end coupled to 

the lock-rod c. At the other end of the rod c is a plate giving 
two indications behind a screen in the block instrument as to 
the state of the lock " free " and " locked." 

Coupled to the upper end of the lock-rod c is an angle 
piece c 2 normally resting on the roller d of the bell-crank d 2 , 

on the other arm of which crank is an armature attracted by 


the magnet e. 

When permission is given by B for a train to leave A 
a current passes through the coils on the magnet so that the 
latter is de-energised and the armature falls, being assisted 
to do so by the spring binding against crank d 2 , and the 
lock rod c, being freed, falls and lifts lock b out of the tap- 
pet a as seen in fig. 85. The indication then shows " free." 

Fig. 83. 

Fig. 84. 

McKenzie and Holland's Lock-and-Block System. 

Fig. 85. tig- 86 - 

The signalman now lowers his signal by pulling over the 
signal lever, and this action raises the lock rod c again, as 
coupled to the tappet a is a lever f which, on being turned 
by the pulling over of the tappet a, reverses the crank b 2 and 
raises the lock rod c. Matters are now as shown in fig. 86, 
and it will be seen that when the lever is put back 
the lock b will fall into the notch a 2 . This 
is placed so as to prevent the signalman put- 
ting his lever fully back and yet sufficiently 
far to restore the signal to danger. The purpose 
of this will be noticed in the description of a third 
rod g. This is the switch rod, and it is raised 
and lowered by means of the lever f, and thereby, 
when the signal is at danger as in fig. 83, con- 
nects the locking coils on magnet e to the line 
wire, and when the signal lever is over, as in fig. 
86, the coils are connected to the contact treadle 
on the line and, through the miniature semaphore 
h to earth, so that when a train goes over the 
electrical contact the coils are again energised 
and the magnet de-energised so that the armature 
on crank d 2 again falls, freeing the locking rod c 
and lifting the lock b out of notch a 2 so that the 
lever may be put fully back. This operation 
causes a second lever / 3 , that is longer than f, to 
turn sufficiently to again raise the lock rod c so 
that the angle piece c 2 again rests on the roller d 
and the connections are again as shown in fig. 83. 



Fig. 88. 
Details of McKenzie and Holland's Lock-and-Block System. 

In fig. 87 are shown details of the instrument, and 
herein will be seen a patented detector to guarantee that the 
armature on crank d 2 shall be properly held. Extra coils / 
are provided in the line circuit to the other coils which, when 
a current passes through, attracts the armature of the lever 
k and thereby withdraws a bolt / from under the armature 
of lever d 2 so that should the latter armature fail to be held 
it cannot fall sufficiently far to free the lock rod. When 
the lever d- is raised, which is done by means of the lifting 
piece m attached to another angle piece c 3 on the lock rod c, 
the lever d 2 passes the bolt I because the latter is bevelled. 

The late Mr. Rollins, of the Great Eastern R., met this 
in another way, which, however, is not so good, as there is 
no guarantee against the lifting piece m being out of adjust- 
ment or the magnet e becoming weakened. In order to 
make sure that the lock rod c had lifted sufficiently to bring 
the armature of lever d 2 in contact with the magnet e Mr. 
Hollins provided a sliding piece n (fig. 88) on the lock rod, 
and this has a slot n 2 cut in it, so that unless it was in line 
the plunger p could not be pushed in. 

The signalman at B sends the releasing current to the 

signal-box in the rear and is prevented, after having 
accepted a train and freed the instrument at A, from again 
using his plunger in the following manner : 

When he accepted the last train and pressed in the 
plunger p (fig. 84) the crank p- was moved by the end of the 
plunger coming into contact with one arm of the crank, the 
other arm of which is placed between the two springs (f<f. 
These springs normally are in contact with each other, but 
they are momentarily separated when the crank p 2 is turned 
and spring q 2 is put in contact with earth, and spring q 3 , 
which is the line terminal, is connected to the battery 
through crank /> 2 , lever p 3 , spring q* and contact q s , the two 
latter coming into contact by the lever /> 3 being moved with 
the crank p-. In plunging, also, the swinger r 2 was pressed 
clear of the rod r and the former and the lock rod r 3 to which 
it is attached fall as far as the contacts qV will allow. To 
the lock rod r 3 is attached a connection to the screen, and 
when r 3 falls it causes " train-accepted " to appear instead 
of " line-clear." 

This improved lever contact prevents the possibility of a 
bad contact being made by plunging very quickly and the 
duration of the contact being insufficient to discharge the 


armature as the contact is made by the contact q 5 falling 
on the contact q 4 . 

As the plunger recedes it withdraws lever p 3 and con- 
sequently contact q 4 comes from under q 5 so that the 
lock rod r 3 drops behind the crank 2 and therefore the 
signalman cannot plunge until the lock rod has been 
raised, which is done by the rod r. This is coupled to 
the crank f (figs. 83-86) and it is raised when the signal 
lever is put back to normal. 

When the signal " train-entering-section " is received 
at B the man there turns the switch hook s so that it en- 
gages with the plunger, which cannot therefore be 
pressed in. This also breaks the line-wire by separating 
the springs q 2 q 3 so that the coil i releases its armature 
t 2 and the indication " train-on " appears in front of 
" train-accepted." When the train has passed and the 
signal lever put to danger, the signalman turns back 
the switch hook 5 so that the coils t are again energised 
and the armature t 2 attracted so that the screen is again 

Tyer's System. 

Tyer & Co., Ltd., have adopted a " Lock-and-Block " 
system to their instrument illustrated by fig. 2. 

The arrangements are shown in fig. 89. 

A train is assumed to be in transit between B and C, 
on the up road, and also one between B and A on the 
down road; both block instruments at A and C showing 
" train-on-line " and the discs locked in that position. 
The discs of the two instruments at B, having been 
cleared by the trains passing into the sections beyond 
the starting signals, show the " train-out-of-section " 
discs, ready to give a like permit for trains to proceed 
from A and C- 

Should C call B for " line-clear," B, in reply, would 
turn his commutating disc to " line-clear " and by press- 
ing in the bell plunger would lower the lower semaphore 
arm of his own instrument and at the same time lower 
the upper semaphore arm of the instrument at C, and 
the platinum contact in connection with the cylinder 
entering the mercurial cup will complete the circuit 
through the insulated cock e, armature /, bed plate g of 
double coil relay, and the electro magnet h, fixed to the 
starting signal lever. The armature of electro magnet 
h being attracted, will break the contact i in connection 
with same and the signal lever can now be pulled over 
and the outdoor signal lowered. 

Upon the train entering the section, C will give B the 
departure or " train-entering-section " signal; B will 
reply by turning his commutator disc, which will show 
" train-on-line," and acknowledge same by pressing in 
his bell plunger. This will raise, the lower semaphore 
arm of his own instrument and at the same time also 
raise the upper arm of the instrument at C, and the 
commutating disc will become locked in the instrument 
at B by being caught upon the step of armature of 
electro magnet d. 



Upon the train arriving at B and entering into the sec- 
tion B-A, the rail contact beyond the starting signal will be 
actuated, thus completing a circuit through the coil / of 
double relay, and the armature / being attracted from the 
insulated cock e on the bed plate g, will break the circuit in 
connection with the signal lever lock h. The armature / will 
be caught up by the hook lever k, fixed to armature I, of coil 
m and the contact springs n, n, in connection with the right 
hand mercurial cup on the dial and the locking magnet d of 
the rear block being closed, the commutator will be released 
to " train-out-o] '-section " upon A acknowledging the depar- 
ture or " train-entering-section " signal. This he will do by 

turning the commutator disc of his instrument to " train-on- 
line " (in which position it will become locked), which will 
raise the lower semaphore arm of his instrument and the 
upper one in the instrument at B. 

The rail contact will also have completed the circuit 
through the signal replacer which will restore the outdoor 
signal to danger. (See fig. 124, page 62.) 

It is impossible for the starting signal to be lowered a 
second time, neither can the rear block be freed until the 
signalman at B has put back his starting signal lever, and the 
man at A has acknowledged the departure signal and the 
semaphore of instrument at B has been raised. 

Fig. 492. Signal "On." Fig. 493. Signal "Off." 

" All-Electric " Signals at Antwerp, Siemens and Halske's System (see page 269). 




In America this method of controlling the block instru- 
ments and out-door signals and interlocking them is known 
as " Manual-Control."* 

Sykes' (American) System. 

On portions of the New York Central and Hudson River 
R. R. ; the New York, Lake Erie and Western, and the New 
York, New Haven and Hartford R. R., Sykes' system is in 

It was introduced in America in 1881, and as the con- 
struction is different to the British pattern already described 
it may be of interest to illustrate it. 

The bell signals are exchanged on a separate instrument 
to that by which a train is accepted and cleared. 

Fig. 90 illustrates the Sykes block-instrument. Passing 
into the instrument are two upright rods, both of which are 
connected with the lever in the locking frame working the 
home or starting signal, or whatever the signal may be that 
is the last stop signal before entering the next section. 

One, a, is called the "plunger-rod." By means of this 
the signalman, when he plunges, locks himself so that he 
cannot plunge again until he has pulled off his signal and 
put it to danger again. The other, c, is called the " lock- 
rod." This rod normally holds the lever fast in the locking 
frame until it is caused to release the lever by an electric 
current from the next box in advance. 

In the position illustrated (fig. 90) the instrument is ready 
for the signalman to plunge to accept a train from the box 
in the rear. On pushing in the plunger b, the upright-rod d 
is, by means of a bell-crank lever, raised and contact is 
formed by the cross-bar e at / and g (see plan). The cross- 
bar e is attached to a rod working in a dash-pot h, and its 
movement is thus retarded and good contacts ensured. 

The contact at / joins up the unlocking circuit to the box 
in the rear and frees the starting signal there, and the con- 
tact at g breaks down this same circuit the unlocking being 
completed in a moment and the circuit is not restored until 
the train has passed through the section and gone over an 
electrical contact at the end of the section. 

The signalman is prevented from plunging a second time 
by means of the trip-rod i which is on the immediate left of 
the plunger, and which has a pawl fe, also a block I. The 
* See footnote 01 page 9- 

plunger-rod a has attached to it a pin m upon which rests 
the pawl fe, so that the plunger-rod a holds up the trip-rod 
i, and the plunger b comes in between. 

The plunger b has on one side a projecting piece n which 
comes into contact with the pawl fe, forces it off the pin m 
and allows the trip-rod i to fall, and the block I to come into 
line with n and prevents the plunger being used. 

Suppose the instrument described to be at signal box B, 
and that the train for which B has plunged is coming from 
A The signalman at B having plunged, is prevented from 
plunging again. The train is travelling from A to B. 
Therefore the next thing to be done is for the signalman at 
B to lower his signal for the train to go forward to C. He 
therefore asks C by speaking telegraph, bell or telephone, 
to unlock his starting signal lever. This lever is normally 
locked by a connection on the lock-rod c, which is held up by 
the pin o on the trip p, having a toothed arm that engages 
in a similarly toothed arm of the lever r connected to the 
armature 5 of the coils t. 

When the signalman at C is in a position to accept the 
train, he presses his plunger and thereby causes the coils t 
to be energised and to attract the armature s, which there- 
fore rises and causes the toothed arm of the lever r to turn 
to the right and the engaging teeth to leave the correspond- 
ing teeth on the arm of the trip p, which, being thus freed, 
falls and draws the pin o out of the slot of the lock-rod 
c which consequently falls by gravity and takes a lock out 
of the starting signal lever which the signalman at B then 
pulls over. 

Pulling over this lever causes the plunger-rod a to travel 
Still further, and the pin m attached thereto (and which got 
on the upper side of the pawl fe when the signalman at B 
plunged to accept a train from A) now forces the pawl aside 
and gets underneath it. Consequently, when the lever is put 
back to danger, the plunger-rod a is raised, carrying with it 
the pin m and the pawl fe, and therefore the trip-rod i, and 
the machine resumes the normal position, except in one 
respect, and that is, that the electrical circuit is broken down 
until the train has really arrived and passed over the elec- 
trical contact, which again restores the connections at signal- 
box A, and allows the coils t to attract the armature 5, 
thereby freeing the lock-rod which holds the starting signal 
lever, so that unless the train has actually passed out of the 



A B section, the signalman at A cannot lower his starting- 
signal, as it is held fast by the lock-rod. 

The electrical contacts used in all systems of " Manual 
Control " in America take the form of a short section of 
' Track-Circuit." 

The starting signal is provided with an electrical slot in 
circuit with the electrical contact, to ensure the signal being 
put to danger regardless of the signalman's action. 

The upper part of the front of the block instrument is 
provided with the two apertures v l v-, behind which appear 
screens indicating the position of the plunger-rod a and the 
lock-rod c. The screen w is attached to the trip-rod i, and 
indicates whether the road is "blocked " or "clear," and 
screen x is attached to the lock-rod c, and shows whether 
the lever is " locked " or " free." 

Colenian's System. 

On several American railways the system invented bj 
Mr. Coleman, of the Union Switch and Signal Co., is in use. 

Fig. 91 illustrates a sectional side view of the block 
instrument in its normal position, and ready to signal a 
train forward. 

When a train has to proceed the signalman pulls out the 
plunger i. To the plunger is attached an angle-piece 2, 

Fig. 90. Sykes' (American) Lock-and-Block. 

which in its travel comes into contact with an arm 3 of the 
crank 4, and moves it so that the top arm 5 of the crank 4 
comes into a vertical position and is held there by the pin 6, 
engaging with the pawl 7 attached to the front of the instru- 
ment, and which is attracted by the armature of the coils 8, 
excited when the spring 9 is forced by the arm 10 of the 
crank 4 into contact with spring n. 

The current on its way to the coils 8 passes through the 
coils of the interlocking mechanism (27 in fig. 92), and the 
circuit is only complete when the lever is in the normal 
position, and the signal at danger, as indicated by fig. 92, 
so that it is ensured that the signalman must have put his 
signals up behind the previous train. 

Another arm 12 of the crank 4 holds up the lock 13, which 
falls when the crank moves so that the hook of the lock rests 
upon the plunger, and when the plunger is returned to the 
normal position by the action of the spring, the hook on the 
lock 13 enters the slot 14 in the plunger, so that it cannot 
be withdrawn again to admit a second train. As the lock 
13 travels, the other end 15 forces the spring 16 against 
spring 17, and so the unlocking circuit is joined up. 

The lever in the signal box in the rear, and working the 
starting signal for entering the 

section, is shown in fig. 92. 

The lever 18 has a lug 19 work- 
ing in a slot of a quadrant 20 of 
the well-known Saxby & Farmer 
type, and this in turn is coupled 
by the link 21 to an arm 22 of the 
block 23. 

The pulling over of the lever 
turns the block 23 to the left, but 
it is not free to do this so long as 
the finger 24 is in the position 
shown in the illustration. This 
finger works on the spindle 25 and 
is raised upwards and free of the 
block 23, when the armature 26 is 
attracted by the coils 27, excited 
by a current flowing through them 
and set up bv the operations at the 
box in advance above described. 

When the road is clear and a train has been accepted by the signalman at 
the box in advance the lever is freed. On the train going over an electrical 
contact fixed in advance of the starting-signal, the coils 27 are reversed and 
the armature 26 is released, and the finger 24 then falls in front of the block 
23 and prevents the lever from being again worked. In case the signalman 
has not put the lever to normal, the finger rests upon the top of the block 
ready to fall into position when the lever is put back, and so long as the finger 
is in this position, the battery is cut out and the signalman cannot accept a 
second train. The result of this is, that the man must have his signal at 
danger before he can accept a train, and having put it to danger, he cannot 
lower it again until the train has been accepted by the box in advance. 

The starting signal is provided with an electrical slot in circuit with the 
electrical contact fixed in advance of it, so that when the train has gone into 
the section, the signal is put automatically to danger, and, as is usual with 



Fig. 92. Interlocking on Lever 

" v "~ 

Fig. 91. Normal Position of Block Instrument. Fig- 93- Block Instrument at 

Union Switch and Signal Company's System. 


electrically slotted signals, it cannot be again pulled off unt'l 
the signalman has put his lever to normal. 

At the end of the section a second contact is fixed, and 
when this is deflected, the coils resume their normal polarity, 
and are again ready to attract the armature. 

When the electrical contact at the box in the rear is 
depressed, the starting signal there is placed automatically 
to danger, as already stated ; also the lever there interlocked 
so that it cannot, for the present, be worked a second time. 
The same current also demagnetizes the coils 8 at the box in 
advance (fig. 91), so that the pawl frees the pin 6, and the 
arm 5 falls. It can, however, only travel as far as the 
position shown in fig. 93. 

There it will be seen that the arm 5 has been stopped in 
its downward travel by the stop 28, on the under side of a 
second arm 29, engaging with the cross-bar 30 on the arm 
5. The arm 29 has hitherto been held up by the coils 31, but 
they become de-energised at the same time as the coils 8. 

On the train passing over the electrical contact at the 
box in advance, and certifying to its passage out of the sec- 
tion, the coils 31 are agitated, the armature 32 of the arm 
29 is attracted so that the arm is lifted, but its upward move- 
ment is only enough to clear the cross-bar 30 from the stop 
28, and immediately another cross-bar 33 engages with a 
stop 34 on the upper side of the arm 29. 

The travel of the crank 4 has not yet been completed, 
although rapidly approaching the normal, but still the crank 
has not travelled sufficiently far for its arm 12 to lift the lock 
13 out of the plunger i. 

This is attained at the next operation, which is that of 
the signalman returning his signal to danger, and thereby 
again freeing the armature 32 from the coils 31, so that the 
arm 29 again falls, and the cross-bar 33 is freed from the 
stop 34. The crank 4 consequently falls, and in so doing 
the arm 12 lifts the lock 13 out of the plunger, and the way 
is again clear for the offer and acceptance of another train. 

The crank 4 also carries a screen 35, which bears three 
indications "free," " train-in-block," and "locked." 
When the apparatus is in its normal position, as seen in fig. 
91, "free " appears at the aperture in the front of the in- 
strument, and when the plunger is withdrawn to admit a 
train, the screen is altered to "locked." Then when the 
train goes over the electrical contact at the box in the rear, 
and the crank 4 falls to the position indicated by fig. 93, 
the " train-in-block " is given. This alters to "free " when 
the train passes out of the section. 

It may also be noted as a matter of important detail that 
the plunger is locked up and cannot be returned to its normal 
position, unless the crank 4 has travelled sufficiently far for 
the arm 5 to be engaged with the pawl 7. This is attained 
by means of the pawl 36, which is provided with a stud 37 
resting normally on the stud 38 attached to the plunger. 
Directly the plunger is withdrawn the pawl 36 falls, and its 
stud 37 comes in the way of the stud 38 on the plunger, but 
when the crank 4 has travelled sufficiently far, the arm 3 
of the crank engages with the pawl 36 and lifts it free. 

In the upper part of the instrument is fixed a miniature 
arm, which is deflected when the coils are agitated. 


Fig. 94. Patenall's Lock-and-Block Instrument 

Fig. 95. Patenall's Instrument. 

Patenall's System. 

Fig. 94 illustrates the Patenall instrument. 

When the signalman wishes to send a train on to the 
next section he pulls out the handle a half-way, and secures 
it in position by putting the knob b between the stops & 1 V-, 
and thereby moves slide c on which are two pins c 1 c 2 . The . 
former, c 1 , engages with the crank c 3 , and the latter, c'-, with 
the crank c 4 . In this operation c 1 simply slides along c 3 , and 
comes into contact with the lip on the top of the same, but 
c 3 carries the arm of c 4 along with it, and the other end, 
being connected with the upright rod d, causes the latter to 
be lowered half its travel. 

At the top end of the upright rod d there is a connection 
with a lever e, at the other end of which are contacts which 
are intended to join other contacts on the plate e l , which has 
two sides. The lower part of one side of the plate e l is joined 
by the wire to the block instrument at the box in advance, 

97- Patenall's Lock-and-Block Instrument. 

and to the other side is joined the wire leading to the magnet 
The lowering of the rod d causes the lever e to rise 
between the two sides of the plate e\ and contact being 
made, the instrument at the box in advance is put in circuit 
with the magnet /. 

At right angles to the upright-rod d, and between it and 
the front of the instrument are two other rods g h, the former 
of which is nearer to the rod d, and in its normal position, as 
seen in fig. 94, it i s kept up by a lug g\ resting on a lug d 1 , 
on a loose lever d*, pivoted at d* on the rod d, whilst ft is 
kept up by its upper part resting on the top of g. As the 
rod d fell partly on the slide c being worked,- the rods g h 
also fell the same distance. 

The signalman now asks the box in advance to send per- 
mission for the train to go forward, and this he does by 
pressing in his plunger t, which comes into contact with the 
loose lever <P on which is a block of insulating material d 4 , 
which forces the spring k into contact with the spring I, and 
so forms circuit with the box in advance. 

Directly the lever d 2 is moved to the right, the lug d 1 
comes from under the lug g 1 and the rod g falls, but the rod 
h does not move for the moment as it is held in position by 
the plunger. The slot in the rod g, through which the 
plunger works, is longer than the slot in h. 

The state of affairs, so far as the plunger is concerned, 
is now as seen in fig. 95, which shows more plainly the 
arrangements in that part of the instrument, and now 
directly the plunger is released from the signalman's pres- 
sure it flies to normal, and the rod h, being freed, it follows 
rod g. 

r. 2 



In the upper part of the front of the instrument there are 
two apertures, and in the lower of these an indicator m 
appears. It is attached to the rod /i which shows a blank 
white face when normal, and this changes to " train-on " 
when the rod h drops. 

The rods d, g, h, and the lever d- are now in the position 
illustrated by fig. 96, and it will be observed that the plunger 
cannot again be used as the aperture is blocked by the rod h. 
If the signalman at the box in advance be in a position 
to accept the train, he plunges, and so sends a current 
through the plate c l and on to the magnet / which attracts 
the armature n 1 of the lever ;/, which carries a latch-lock n- 
fitted in the slots c b c of the slide r. When the slide is 
"home," the latch-lock is in the slot c 5 which is wide, 
enough to allow the slide to travel half-way with the latch 
in the slot. 

The latch-lock being lifted, the slide c can be pulled fully 
out, and the pin c 1 moves the crank c 3 so that it lifts the rod 
o. This rod passes down into the locking frame, and when 
ir is lifted unlocks the lever for entering the section. On the 
top of the slide c is a projection p which passes under a 
roller p l of a lever p- connected to the indicator r fixed in the 
upper aperture of the instrument, and which now changes 
from " clear " to " blocked." 

The pin c" on the slide c also moves the crank c* so that 
the upright rod d falls the remainder of its travel, and in so 
doing the loose lever J 2 is forced to the right in order that 
the lug d 1 may pass the lug g l , and then the former gets 
under the latter. The downward movement of the rod d also 
raises the contacts on the lever e to the upper part of the 
plate f 1 and joins the magnet / up with the " Track-Circuit." 

The passage from one condition to another demagnetizes 
the magnet / when the armature n 1 falls away and the lever 

n allows the lock ir to enter the slot c 6 , and the slide is locked 
and the instrument is in the position seen in fig. 97. 

Immediately the train clears the section and has passed 
oti the " Track-Circuit " the magnet / is again energised 
and the latch-lock 2 lifted out of the slide. It can be put 
back, and in so doing the rod d is raised, the lever c 
lowered, and the rods g h raised, so that the instrument is 
restored to its normal condition. 

Other references are made to manually-controlled sys- 
tems of signalling in America in Chapter X. dealing with the 
methods of working single lines. 

Winter's System. 

In Chapter X. there is a description (figs. 224-227) of 
Winter's instrument, which is a form of " Lock-and-Block " 
used on Indian railways. 

McKcnsic and Holland's System for Indian Railways. 

This is an arrangement modified from that illustrated in 
Chapter III. (figs. 83-88; p. 42) and which only requires 
one wire. 

Fig. 98 is a diagram showing two stations A and B and 
the fixed signals thereat. In order to understand the system 
let the operations be followed for the passage of a train 
from A to B- 

No. 4 starting signal at A is electrically controlled from 
signal box B. It is held locked by the lock a resting in the 
tappet of the lever. 

When A is read}' to send the train to B he presses his 
bell plunger B 1 which sends a negative current to the relay 
at B so attracting armature b to the contact point />'-, thereby 
picking up a positive current and sending it through coils 
B- and ringing the bell. B replies and presses in his plunger 
c, which allows the screen d to fall and to indicate train- 

Fig. 98. McKenzie & Holland's System of Lock-and-Block for Indian Railways. 


accepted and joining up the line springs with the battery 
contacts e and, picking up a positive current, sends it 
through the line to the relay at A and putting over armature 
/ to contact point / 4 , thereby picking up a positive current 
and sending it through the locking coils g so that the arma- 
ture g- falls and roller g* is moved from under the connec- 
tion on down-rod a- whereby the lock a is lifted out of the 
tappet, and the screen is changed from locked to free. 

When B pressed in his plunger he put his plunger lock 
P L in circuit with the electrical treadle T at B and at the 

g. qq. l r ig. Io ' F'g- IO1 - 

McKpnzie & Holland's (Indian) Lock-and-Block. 

same time the plunger becomes locked as shown in figs. 99, 
100 and 101, and it cannot again be pressed until the train has 
gone over treadle T, which causes the coils P L to attract 
the armature c 2 . 

Part of this arrangement is the provision of Deakin's 
mechanical replacer (fig. 203 in Mechanical Railway Signal- 
ling, p. 84), marked M.T. on fig. 98 for automatically 
restoring the starting signal at A. When the signalman puts 
back his lever the cheese-cutter lock a 3 raises the rod a- and 
this allows the lock a to again fall into the tappet and the 
lever is again locked. 

Siemens-Halske 's System. 

The system of Siemens & Halske, of Berlin and Vienna, 
is in use throughout the whole of the German railways and 
on many other Continental lines, including i,fi^ kiloms. 
(1,015 miles) of the Belgian State railways. 

The principle is that before a signal can be lowered for 
a train to enter a section it must be accepted by the man at 
the advance station unlocking the lever working the signal. 
Also that before a block signal can be given to the station 
in the rear indicating that a train has arrived the train itself 
must have gone over a section of " Track-Circuit " in ad- 
vance of the home signal. 

Fig. 102 is a view and fig. 103 is a drawing, partly in 
section, of the apparatus as fixed in the signal-box or sta- 
tion. The up and down home signals are respectively 
coupled by chains to the levers a a-, which are turned in a 
downward direction through half a revolution from back to 

The unlocking mechanism is illustrated by the diagrams 
figs. 104 to 107. 

Fig. 104 shows the signal " free," with the lock b 
taken out of the cam a on the signal lever, and this is the 
normal condition. When a train has to be sent, the signal- 
man presses down the plunger c, which carries with it the 
rod d, also the second rod d-. This puts the lock b into the 

lever , but should the lever not be in its proper position the 
lock b will not fall sufficiently to allow the plunger and the 
rods d, d- to travel the distance required. If, however, the 
lever a is in its proper position the lock b will enter; also, 
as a consequence, the contact spring c 3 carried by the lever 
c, which is pivotted and the head of which e" rests under 
the rod d, will make contact with e* and the spring e 5 will 
break contact at e 6 . 

Matters are now as shown in fig. 105, and a circuit is 
completed through the magneto-generator f to earth and 
allows the signalman to send a current to the station in 
advance asking for permission to send a train. This rings a 
bell in that box, and providing the previous train has arrived 
and passed over the section of "Track-Circuit," the man 
there can press down his plunger g and then, by turning his 
magneto-generator, sends an alternating current through the 
coil h, so attracting the armature; of the anchor/ 2 , whereby 
the alternating current gives an oscillating movement to the 
same and allowing the screen fe to fall as seen in fig. 106. 

The lock b is now fastened in, as although the signalman 
releases his hold of plunger c and it rises and takes with it 
the rod d, owing to the spring J 3 (as seen in fig. 107) the rod 
d- cannot rise as the latch d 4 at the top of it rests under the 
pawl t/ 5 and this pawl is held by half of the head of the axle 
k- projecting and preventing the pawl returning, although 
drawn by the spring d 6 , until the screen k has again been 
raised. When, therefore, the train arrives at the box in 
advance the man there operates his magneto-generator /, 
which again energises the coil h and attracts the armature 
;' and the anchor f 2 , so freeing the screen k. The frame d 7 , 
which acts as a guide to the rod d, and which does not rise 
with the latter, as it the guide d 7 is held by the pin fe 3 on 
the screen fe. When the latter is, however, freed, the spring 
</ E under guide d 7 forces the latter upwards and carries with 
it the screen k. As soon as the projecting half k- is hori- 
zontal the pawl </ 5 is pulled back by its spring and this frees 
the rod il-, which is pushed up by its spring so that the lock 
/> is freed and lifted out ot the wheel a by the spring attached 
to the lock. 

When the rod d- rises it pushes away the pawl I that has 
been forced under the foot of the rod d so as to prevent the 
signalman again pressing his plunger. 

Indicators m m (fig. 103) are provided to show when per- 
mission has been given to free the signals at the adjoining 
stations and indicators n are for showing when the signals 
are freed by the other boxes. Indicators a o are for inti- 
mating when a train has passed over the Track-Circuit and 
allows for the train-arrived block signal to be sent. 
Fig. 1 08 is a diagram of the electrical connections. 
Assuming that a train is travelling from left to right on 
the lower line and is approaching the second signal box from 
the left. The distant signal on the left of the arrow and the 
stop signal on the right would be "off," and the block instru- 
ments showing the line occupied. When the stop signal is 
lowered a switch (the upper of the three to the left of the 
signal box on the diagram) would join up a local battery to the 
electro-magnet and thence to the rail contact seen just to the 
right of the signal. The two lower switches are coupled, 



m n n m, 

Fig. 103. 

Siemens and Halske's Lock-and-Block System. 

Fiir. 102. 

Fig. 104. 

Fig. 105. 

Fig. i 06. 

Fig. 107. 


108. Diagram of Electrical Connections, Siemens and Halske's Lock-and-BIock. 

one to each rail of the insulated portion, and when a train 
passes over this " Track-Circuit " the lock in the signal box 
could be withdrawn as the armature of the electro-magnet 
seen in the top left-hand corner of the connections in the 
signal would be attracted. In order, however, to prevent 
this being done by the first wheel of the train, the rail contact 
is put in to act as a cut-out and prevent the armature being 
attracted, but when the contact is no longer depressed that 
is when the last wheel has passed off it the circuit is com- 
pleted and the lock withdrawn. 

Further reference to this point is made in the description 
of fig. 121. 

Sarroste and Loppe's System. 

This is used on the French State Railways, and under it 
no train may enter a section unless it has been accepted by 
the box in advance, which acceptance cannot be given until 
the preceding train has gone over an electrical contact and 
the stop signal has been placed to danger behind that train. 
The acceptance of the train by the box in advance unlocks 
the lever working the signal for entering the section, which 
signal can subsequently be put to danger at any time, but 
cannot again be pulled " off " until another train has been 
accepted by the box in advance. This is done by the usual 

The novelty of the system is the block instrument, which 
is illustrated by fig. 109. 

On the upper part of the instrument is a dial by which the 
signalman sends messages to the box in advance. On the 
dial in the inner circle are numerals from i to 12, and in the 
outer circle are discs lettered as follows : 

1. Blank. 

2. Call attention. 

3. Speak. 

4. Train offered. 

5. Train refused. 

6. Train accepted. 

7. Train left. 

8. Cancel message. 

9. Testing signal. 
10 and ii. Blank. 

12. Train permitted to enter 
occupied block. 

In the centre of the instrument there is on the left-hand 
side a similar dial but smaller, and upon this the signalman 
receives messages from the box in advance, and on the right 
of this is a push button ringing a similar bell to that seen in 
the top part of the instrument in the side elevation of fig. 

When the man wishes to send a message he presses the 
button, and turns the pointer to No. 2, and in reply his 
receiving dial pointer turns to No. 3. When the message is 
completed the pointer has to be returned to its normal 
(middle) position. 

In the bottom part of the instrument is the switch which 

Fig. 109. Sarroste and Lopp^'s Lock-and-BIock. 

works to three positions a, b, c. In fig. 109 the switch 
stands at " is-line-clear," when turned to b it indicates 
" line-clear -g'vv en," and at c it shows "line-clear-obtained." 
The movements of the switch are governed by the electrical 



Fig. no. Cardani's System of Lock-and-Block. 

treadles on the line, which in turn govern the lowering of 
the outside signals. 

By the side of the switch is an aperture d, behind which 
an indicator appears when a train is in the section. 
Cardani System. 

This is a hydro-electric method in use on the State Rail- 
ways of Italy. It is manufactured in Italy by Gio. Servettaz, 
of Savona, and in this country by Saxby & Farmer, Ltd. 

The block instrument has a handle the normal position of 
which is in the centre and which turns to the right to indi- 
cate "clear," then round from right to left for " train-is-in- 
section," and the circle is completed for "train-hus-arrived." 

A hydro-electric slot (fig. 128, p. (14) is provided on the 
starting signal, the principle of which is that an electric 
current is sent into the slot when the signalman at the ad- 
vance box accepts the train. This closes a valve in the slot 
so that when the signalman works the lever the liquid, com- 
posed of glycerine and water, forces up the signal-rod and 
lowers the signal. If the train be not accepted the valve 
remains open and the action of the signalman simply forces 
the liquid into another chamber. 

A hydro-electric treadle (fig. 120, p. 61) is fixed in ad- 

vance of the starting signal whereby the compression of the 
liquid in one chamber forces up a piston on the end of which 
is a contact piece which, coming in touch with two contact 
springs, joins them and forms an electrical circuit whereby 
the signalman is free to send notification that the train has 

In fig. no is a diagram of the wiring between two signal 

The instruments working between two boxes A B are 
shown. On the left the electrical connections are as when 
normal and on the right as they are joined up when A is asking 
B for permission to send a train. The upper needle a indi- 
cates, when to the left, that the section is unoccupied, and 
when to the right that it is blocked, and the lower needle 
b shows, when to the right, that consent has been given to 
the box in the rear to send a train. Bell signals, given by 
the plunger />, arc exchanged before any indications are made, 
and the handle HI indicates the three positions normal, con- 
sent, blocked. This also operates a lock on the lever / work- 
ing the signal .S' 1 . Behind the screen d appears an indicator 
coloured green when the section is clear and one coloured red 
when the section is occupied. 



When a train has to be sent from A to B the man at the 
former post presses in his plung-er p and this joins up contacts 
10, 13, and n, 14 at A and energises the magnets E 2 in B, 
attracting a 1 and joining up 15, 16 so that the two magnets 
E 3 are energised, attracting the armature cr which rings the 
bell. The man at B responds to the signal, and if he can 
accept the train he turns his handle m to consent and at the 
same time signifies to himself, by the needle b being turned to 
the right, that he has given permission for the section to be 
occupied. When plunging now, B joins up 10, n and 13, 14 
and a current flowing to A energises the magnets E 2 in that 
box, attracting a 1 so that the bell rings there. So long as 
a 1 is not attracted it holds the anchor-shaped piece a 3 , but 
when a 1 is attracted the latter is free and the current to E 2 
passing through magnet E 1 the latter is energised and turns 
u 3 , the contact of which joins up 17, 18, and thus completes 
the circuit to magnet M of the slot on signal S 1 , so that when 
the signalman moves the lever I the contacts 19, 20 are joined 
up and current sent to magnet M so that the signal may be 
lowered. (See fig. 128.) 

In the event of the signalman at A omitting to put the 
signal S 1 to danger behind a train this would be done auto- 
matically by the train passing over the treadle P 1 (see fig. 120), 
which would join up the contacts, energising the two magnets 
E 5 and attracting a 4 , the other end of which would come 
against the point u> of the anchor a 3 so that the latter would be 
turned and contact broken at 17, 18, thereby de-energising the 
slot M and allowing the signal to go to danger. The anchor 
would not be turned fully to normal and the needle a would 
only go to the mid-position. 

The man at B would, on being advised of the entrance of 
the train, lower his signal S 2 and turn the handle m of his in- 
strument from consent to blocked. He should at the same 
time cancel the permission given to A for the train, and with 
the handle m in its new position the pressing of the plunger 
p would cause the magnets E 1 and E 5 at A to be de-energised 
so that the anchor a 3 and the lever a 4 go to normal. 

A plunger is also provided to release the signalman in 
case he locks himself up or should the instrument fail, but the 
treadle P 2 has to be first depressed and the coils E 6 energised, 
guaranteeing that the train is out of the section. 

Fig. 469. Bridge of " Crewe " All-Electric Signals at Severus Junction, X.E.R. See page 260 



THE term contact maker covers all forms of treadle, but 
the term treadle does not cover all forms of contact makers. 

There are numerous forms of treadles used in associa- 
tion with " Lock-and-Block " and other purposes in signal- 
ling, some of the mercurial type and others with contacts. 

They have various uses. Some lock the levers in the 
locking-frame and others unlock them, and whilst this is 
effected in most cases by the deflection of the rail acting on 
the short end of a lever, it is also attained by the depression 
of a bar, attached to the rail, similar to a facing point locking 
bar. This latter kind is, however, generally used as a fouling 
or clearance bar. 

Sykes' Rubbing Contact. 

This is of the bar type, and its purpose is to keep certain 
signals locked at danger when the road is not clear, owing 
to vehicles standing in such a position as to be foul of the 
line, upon which a train has to travel. This is of the rubbing 
contact form and is shown by fig. in. 

Attached to the sleeper is a bracket a on which works on 
one side the bar b, and on the other side a heavy balanced 

Fig. in. Sykes" Rubbing Contact Treadle. 

rod c to which is attached another rod d which passes under 
the rail and actuates the contact maker e. If a vehicle be 
standing foul, the bar b is depressed and the rod c rises and 
so draws the rod d, and the switch / is taken out of contact 
with the insulated piece g and the circuit is broken down. 

Sykes' Double Contact. 

This is of the mercurial type and is illustrated by fig. 112. 

It is carried by two bolts a 1 a 2 secured to the rail, and has 
two levers b 1 b"- attached at c 1 c 2 . The short ends of the 
levers rest under a pin d secured to a third bolt a 3 midway 
between the other bolts. The other ends of the levers b 1 b 2 
rest under cups of mercury e, in which are wire connections 
to earth and to the lock on the signal lever. On an engine or 
vehicle passing over the rail, the depression bends down the 
pin d and raises the long ends of the levers b 1 b 2 , and so the 
cups of mercury are tilted and contact is made or broken. 

&a ~ 

L" X ~ 


a 1 

^ . _^_ _^^ ^s 



1 1- -1 


Fig. 112. Sykes' Double Contact Treadle. 

In the large scale detail the cup of mercury c is shown 
with the two contacts f 1 f. By the depression of the short 
end of the lever b 1 , the cup of mercury is tilted by the raising 
of the long end of the lever. 

Hodgson's Treadle. 

This is shown by fig. 113, and is also of a mercurial type. 
Attached to the rail by a clip, and fixed midway between 

Fig. 113. Hodgson's Treadle. 

two sleepers, is one end of the lever a pivotted on the pin b, 
and to the other end of the lever is a short rod c connecting a 
box d containing mercury, and which is carried on a lever 
pivotted at e. Through the box d passes the wire for elec- 
trically locking or unlocking the signal, and this wire is 
carried over the mercury and normally free of it. 



Fig. 114. Siemen's Hydrostatic Contact Maker. 

On the rail being depressed by the passage of a train, the 
short end of the lever a is deflected and the long end raised, 
and so, by means of the short rod c, the box d is tilted from 
left to right, and the mercury being disturbed comes into 
contact with the wire, and the circuit is completed or broken 

Siemens Bros. ' Hydrostatic Contact Maker. 

This is secured to the rail, fig. 114, by four bolts, two of 
which, a 1 a 2 , are shown in the elevation. Under the rail is a 
piston b, the bottom part of which rests on the diaphragm c. 
Round the top of the piston there is a ring of indiarubber to 
prevent dirt getting in. 

In the bottom of the apparatus is a supply of mercury 
which passes up the pipe d into the cup e, which has inside 
a five prong fork, the upper end of which is carried by the 
casting /. An electric wire enters the treadle through the 

pipe g, and is connected to the casting /, which casting is 
insulated from the rest of the instrument by a slate cap h. 

On the piston b being depressed by a train passing over 
the rail, some of the mercury is forced up the pipe d, and 
joins up the contacts of the five prong fork.. 

Stevens and Sons' Treadle. 

The apparatus is carried in a box under the rail as shown 
in fig. 115, and held in position by a wrought-iron strap 
passing under the two adjacent sleepers and each end secured 
to the rail. 

Bolted to the rail is a bracket a which carries a plunger b, 
which passes into an air-tight box. On the rail being de- 
pressed the plunger is deflected and raises the longer end 
of lever d which acts upon the other lever d, which 
acts on the lever e, at the end of which is a platinum point 
that comes into contact with the spring /. To this spring is 
connected the line wire which enters the box along with the 
other wire (connected to the lever c) at g. When the lever e 
comes in contact with the spring / the circuit is completed. 
Rollins' Treadle. 

The late Mr. Hollins, when electrical engineer of the 
Great Eastern R., designed a rubbing contact treadle which 
is illustrated by fig. 116. 

It is contained in a box by the side of the line and con- 
sists of the lever a, pivotted at b, the short end c of which is 
let into a corresponding hole in the rail, and the longer end, 

Fig. 115. Stevens and Sons' Contact Treadle. 

Fig. 116. Hollins' Contact Treadle. 

which is weighted at d, has attached to it by means of steel 
bar e a friction clutch / carrying the contact piece g, and 
moving up and down between the spring h l 7i 2 (see end view). 
The novelty of this form of treadle lies in the clutch /, 
which is raised on the rail being depressed ; but in case the 
lever gives an extra big sweep the clutch cannot travel 
further than between the stops j 1 f, the remainder of the 
sweep of the lever being met by the steel bar e sliding 
through the clutch, so that the required movement, and no 
more, is always obtained. 

Mercier's Treadle. 

This can be clipped to the rail or bolted through as shown 
in fig. 117, and it is provided with adjustable rubbing con- 



tacts. The wires are in a cable which passes in through a 
watertight gland. A strong circular rubber ring is used to 
make the lid-joint watertight, and a rubber cap is fastened 

ftfuMer Join t Ring 

*" _ rrn rrn 


f 'acted Gland 


Fig. 117. Mercier's Treadle. 

to the case the lever works in, the whole forming an air and 
watertight chamber. 

Tyer's Treadle. 

This is of the rubbing contact type, and is shown in fig. 
118. When a passing train deflects the rail the arm is 
depressed, and this, acting on the short end of the lever b, 

McKensie and Holland's Treadle. 

This, fig. jig, is also of the rubbing contact type just 
described. Secured to the rail by the brackets 2 2 is a frame 
containing two levers 22 22 which are pivotted at 24 24 and 
the short ends of which rest under the end of an arm 23 
which is also attached to the rail. 

When the rail is depressed by the passage of an engine 
or train the arm 23 is deflected so that the longer ends of the 
levers 22 22 are raised, which operate the arms 19 19 and 
thereby press in the knobs 18 18 so that an electrical circuit 
is completed owing to springs 5 6 being put in contact. 
Cardnni's Hydro-Electric Ctnitucl Maker. 

This is also a device which is operated by the deflection 
of the rail. 

In a casting by the side of the line is a two-cylinder 
chamber, fig. 120, containing a liquid composed of glycerine 
and water. When the rail R is depressed an oscillating 

Fig. nS. Tyer's Treadle. 

n C 

] I I 


I 1 ] 1: ; 





turns, to the left, the lever c of the commutator d so that con- 
tact between the spring e and plate / is broken and is made 
with plate f. 

movement is given by the lever B, attached to the rail R, 
to the piston P, which causes the liquid in the chamber A 
to be driven into the smaller chamber a and the piston p to 
be forced up until its end c makes contact with the springs 
r 1 r. The two chambers communicate with each other by 

Fig. 119. McKenzie & Holland's Treadle. 



means of the ports D d which are forced open by the flow of 
the liquid against their pointed ends. The port C is always 

Kig. 1 20. ("urclani's Hvdro-Electric Contact Maker. 

open to let the liquid that is above the pistons pass from one 
chamber to the other. 

Position of Contact-Maker. 

The difficulty mentioned in Chapter II. as to the posi- 
tion of the treadle for releasing the block instruments in 
Lock-and-Block, viz. , that the contact maker or treadle must 
be placed well in advance of the signal in order to cover a 
long train, and thereby be too far ahead for a short train, 
might, the Author thinks, be obviated by placing a locking 
bar similar to that illustrated by fig. i 1 1 in the rear of the 
contact maker, i.e., before the contact maker is reached. 
The circuit from the contact maker would then pass through 
a contact plate connected with the locking bar and the cir- 
cuit would be broken when the bar was depressed and the 
unlocking circuit would not be completed until the bar was 

Two advantages would be achieved by such an arrange- 
ment. Firstly, the line would not be cleared until all the 
train had passed, and, secondly, a long train would be served 
as efficiently as a short train regardless of the position of the 
contact maker. 

Or this end might be attained by the use of a section of 
" Track-Circuit " as is done by Siemens Bros. 

As soon as the line is set for the passage of a train a cur- 
rent Hows from battery I, fig. 121, through electro-magnet 
2 and lead 3 to the rail contact 4 and through earth to the 
earthed pole of the battery, the armature of the electro- 
magnet 2 being thereby made to keep a contact 5 open. 
When a train passes over the insulated rail 10 and opens the 
contact 4 the electro-magnet loses its current so that its 
armature in falling off closes the contact 5, and thereby 
closes the circuit : battery i, electro-magnet 6, contact ^, 
lead 9, insulated rail 10, train axle, earth, and back to the 

battery. The circuit of the 
electro-magnet 2 remains inter- 
rupted at the contact 4. Electro- 
magnet () being energised at- 
tracts its armature, thereby 
closing contact 7, which by 
means of lead 1 1 closes the cir- 
cuit of the electro-magnet 8, 
whose winding is connected to 
earth, and which represents 
the safety appliance to be 
Fig. 121. actuated. The current which now 

passes through this magnet, which is in parallel with the 
connection through the train axle to earth, is, however, not 
of sufficient strength to excite it so as to influence the 
safety appliance ; it is only when the last train axlt has left 
the insulated rail that electro-magnet 8 receives a full cur- 
rent from the battery through electro-magnet 6, contact =5, 
leads 9 and 1 1 and contact 7 to earth, and this actuates the 
safety appliance. 


Replacers are sometimes called " Reversers," and in 
America are known as " Slots." They are appliances which 
permit a signal to go to the " on " position automatically by 
the passage of the train and quite independently of the sig- 
nalman, also for controlling a signal direct and not through 
its lever in the locking frame. In most systems of " Lock- 
and-Block " such an appliance is provided. 

The American term " slot " applies equally well because 
by the use of a " slot " a signalman cannot lower his signal, 

if it be controlled from another box, until the magnet 

where electrical reversers are used has been energised by 
the action of the signalman in advance. Where "Track- 
Circuit " is employed and a mechanically operated signal is 
controlled by the section in advance one of the replacers 
described below is employed. 

Sykes' Replacer. 

The upright rod on the signal post, fig. 122, connecting 
the signal arm to the balance lever at the foot of the post 
to which the signal wire is attached, is in two parts. To 
the upper part (working the arm) is attached at a a case into 
the lower part of which enters the portion b of the upright 
rod that is secured to the balance lever coupled to the wire 
from the signal box. Upon the rod b rests the weight c 
attached to the arm d. The arm is secured to the casting 
and is pivotted at d 1 . The arm d is held in the normal posi- 
tion by the catch e, to which is attached a rod /. 

On the signalman pulling over his lever he raises the rod 
b, and this bearing on the weight c raises it, and along with 
a : I d 

Fig. 122. Sykes' Signal Replacer. 



it the arm d and consequently all the casting, and therefore 
the rod a, and so the signal arm is lowered. The apparatus 
also contains a pair of coils g in circuit with an electrical 
treadle, the armature h being connected to a catch j, which 
holds up a hammer k. 

On a train going over the electrical treadle the armature 
h is attracted, raises the catch j, and the hammer fe is re- 
leased, which falls on the rod /, consequently depressing 
momentarily the catch e. This frees the arm d, and the top 
of the rod b being bevelled, the weight c rolls off and down 
comes the replacer, restoring the signal to danger. As the 
arm d swings to the right it raises the hammer k so as to 
again engage with the catch e. 

Until the signalman puts back his lever the rod b remains 
inside the replacer, as seen in the right hand illustration, 
and when the rod b goes down, the arm d resumes the normal 
position, as seen in the left hand illustration, and the signal 
can again be worked. 

Another slot of the Sykes Co. is illustrated by fig. 123. 
It is of the divided upright rod type. The balance lever is 
attached to rod B 1 and the signal arm to rod B on which is 
carried the casting b containing the slot. To the rod B 1 is 
connected a thrust rod / sliding in the casting b 1 . This, 

Fig. 123. Sykes* Electrical Rr>p!acer. 

when raised, acts against the stop G on the lever J, the 
lower end of which is pivotted at g and the upper end en- 
gages against the end of the trigger K. This trigger is 
centred at k and one end is connected to the armature h of 
the magnet H. If the magnet be energised and the arma- 
ture attracted, the free end of the trigger K would be low- 
ered so that the lever J would be held when the rod B 1 were 
raised and so the signal would be lowered. If, however, the 
magnet were not energised the lever J would be forced 
over to the dotted position J 1 and the rod B, to the signal 
arm, would be inoperative. 

The weight ;' assists in bringing the lever J back to 


Tyer's Replacer. 

Fig. 124 shows an electrical replacer working on the 
negative system, which requires a current to restore the 
signal after it has been lowered. When the electro-magnet 

Fig. 124. Tyer's Electrical Rpplacer. 

p is not energised by a current, the armature q falls away 
from the poles of same, leaving the lever r locked by the 
slotted drum s, and thereby allows the signal to be lowered; 
directly, however, a local circuit is completed, say, by the 
train passing over a treadle, the electro-magnet is energised, 
the armature q is attracted, the drum s is turned by the 
lever (, attached to armature q, the lever r becomes unlocked 
and the two vertical sliding bars M, l become disengaged, 
the bar u ', which is connected to the crank arm of semaphore, 
is thus freed from M, and the semaphore immediately flies to 

The whole arrangement is mounted upon a plate v, which 
is securely fixed to the vertical bar u. Two brass connection 
plates, w, ia l , arc fixed to the back of case and connected to 
the electro-magnet by the stranded contact wires x, x 1 . 

O'Donnell's Replacer. 

This replacer is illustrated by fig. 125. On the spindle 
of the signal-arm is fixed a circular box with a lug 15 at the 
side to which the upright signal rod is attached instead of to 
the arm casting. On the spindle is also fixed a disc 12, and 
the box is relatively mounted on the disc, and an electric 
clutch device is also mounted thereon. This consists of an. 





Fig. 125. O'Donnell's Rotary Slot. 

electro-magnet 16, carried on bracket 17, and a pawl 19 
which fits into a slot 20 in a recess in the box and is held 
there by the crank 22. Should the magnet 16 not be ener- 
gised the pawl 19 would be forced out of the recess when 
the signal rod connected at 15 was raised, and therefore 
the box would revolve without turning the disc or, conse- 
quently, the spindle. But if the magnet be energised the 
armature 27 of the lever 23 would be attracted. This lever 
works on the same spindle 25^ as the shorter lever 24, which 
has a lug 240 engaging with the crank 22 so that the magnet, 
being energised, causes crank 22 to be held, and this keeps 
the pawl 19 in the recess of the box so that the latter, being 
turned, causes the disc 12 and the spindle to be turned and 
the arm lowered. 

When the magnet is de-energised the pawl is released 
and the box returns to normal by its own weight. 

Hall's Replacer. 

Fig. 126 illustrates the Hall Signal Co. 's replacer. The 
apparatus is carried in a box attached to the signal post by 
means of four screws through the holes a a a a. The upright 
rod of the signal is in two parts, the upper part b being 
attached to the arm and the lower, c, to the rod or wire from 
the signal-box. The lower end of rod b and the upper end of 
rod c- overlap each other, and cut in both is a notch d, into 
which enters a projection from the latch e that is pivotted 
at / to a sleeve g rivetted to the lower rod c. To the box is 
secured a powerful magnet h, and on the lever / is an arma- 
ture k. This armature is secured by a threaded pin with a 
semi-spherical head, which is so shaped as to cause good 
contact to be made should the lever ;' be out of adjustment. 
The armature k is kept close to the magnet h by the compres- 
sion of the phosphor bronze spring /, and the lever ; holds 
the projection on latch e in notch d so that when the lower 
rod c is raised, should the magnet h be energised, both rods 
are raised and the signal pulled "off." If, however, the 
magnet is not energised the lower rod c, owing to the shape 
of the notch d, forces the latch e away, and the upper part b 
is not raised nor, consequently, is the signal lowered. 

Let it now be assumed that all has been in order and the 
signal lowered and that, a train having passed, the signal 
has to be restored, or it may be imagined that for some other 
reason the signal has to be put to danger independently of 

Fig. 126. Hall's Replacer. 

the signalman. In that event the magnet h would be de-ener- 
gised, and consequently the lever / would no longer retain 
its hold on the latch e, so that the weight of the spectacle 
and the rod itself would force down the upper rod b so that it 
would force out the notch d and would go to normal, but the 
rod c would remain up until restored by the signalman owing 
to the pin m working in a slot n in the lower rod. A roller o 
is provided on the latch e to allow for the free movement of 
the upper rod b. 

A dash pot p is attached to the upper rod c so as to 
cushion the effect of the signal being thrown to the " on " 

When the lower rod c is restored the projection on latch e 
again enters the slots d, so that the armature k on lever ;' is 
again in contact with the magnet h. 

As illustrated the door of the box is open, and the latch q 
is used to secure it when closed. 

When two slots are used, as for two separate signals on 
one post, a double slot in one box is provided. A single slot 
weighs 85 !bs. and a double slot 145 Ibs. 



Union Switch & Signal Co.'s Controller and Replaccr. 

In fig. 127 is shown the details of this appliance. The 
upright-rod is divided in the middle. The upper portion 
a is connected to the signal arm and the lower portion /; is 
coupled to the balance lever and is raised and lowered by the 
action of the signalman working the lever. It should be 
remembered that in America all stop signals are worked by 
rodding and not by wire, and therefore a positive up and 
down motion is given to the rod b, and it does 'not fall by 
gravity as would be the case in this country. The ends of 
the rods a b pass into the case c which is attached to the 

signal post and to the ends is attached at d" a frame d. 
Carried by the frame d is a magnet c attracting an armature 
/ pivotted at g. In the frame d is a lever h which has a roller 
/ normally resting against the spring k attached to the lower 
side of the case c. Coupled to the rod a at I is the pawl m 
and between the pawl and the top of the lever h is the roller 
n. To the left of the pawl is the crank o, the short arm of 
which is held down by the armature / when attracted by the 
magnet e. 

Should the magnet be energised to allow the signal to be 
lowered the armature / would be attracted, the crank o would 
be held, and consequently the rod b, rising, would raise lever 
h. which would force roller n upwards and that would also 
raise the pawl m and consequently the upper rod a. When 
the train passes over the relay on the line the magnet e would 
be de-energised so that the long arm of the crank o would 
move to the left, release the roller n and the upper rod a 
would fall by its own weight and the signal go to danger. 

Should the magnet not be energised when the signalman 
pulls over his lever the lever li would, on being forced up- 
ward, push away the roller n, as the latter could not be held 
by the crank o, as its short arm would rise owing to not being 
kept down by the armature /. 

The upper portion of the mechanism is constructed with 
a pneumatic dash-pot p to relieve the shock when the signal 
goes to danger. It is provided with a fixed roller guide r 
inside the cylinder. 

Kig. 127. Union Switch Signal Co.'s Controller and Replacer. 

Fig. 128. Cardani's Hydro-Eleclric Slot. 



Cardani's Hydro-Electric Slot. 

Fig. 128 illustrates a hydro-electric slot in use on the 
State Railways of Italy. It is fixed on starting signals and 
has two chambers A C containing a mixture of glycerine and 
water. The port D, giving communication between the 
chambers, is normally open. - \Yhen the signal may be 
lowered the magnets MM are energised and attract the 
armature F. Weight G therefore lifts the lever E and the 
rod H which closes the port D. Consequently when the 
signal lever is pulled the upright rod T is raised and forces 
against the liquid in chamber A. Were the magnets M M 
not energised and the weight F lifted the liquid in chamber 
A would be forced into chamber C, but as the port D is 
closed this cannot be done, and, as a consequence, the whole 
of the slot and the signal rod B is raised and the signal 
lowered. As soon as the magnets M M are de-energised, as 
by a train going over the treadle, the armature F falls and 
opens the port D so that the weight of the signal rod forces 
the liquid out of A into C and the signal rod B falls so that 
the signal goes to danger. 

Mechanical Replacers. 

For restoring signals by the passage of a train replacers 
are occasionally provided which are mechanically operated. 
Two of these Tver's and Deakin's are described in 
Mechanical Rail-way Signalling, pp. 83 and 84. 


The detecting of facing or trailing points by electricity is 
incomparably better than doing the work mechanically, as is 
the general custom. The Author has had considerable ex- 
perience in this respect. Except in very simple cases where 
the detector is worked by rodding or by wire, considerable 
trouble is introduced in the working of signals, and this 
is greatly increased where more than one set of switches is 
concerned. The adjustment of the signals becomes a prob- 
lem, whilst the additional weight to be moved makes the 
signals hard to pull off. There is the further objection of 
the increased number of rods and wires that are introduced, 
which have to be laid down and frequently boxed in, and 
over which the shunting staff are apt to stumble. Conse- 
quently the use of electrical detectors is to be commended, 
as they do the work better and are no more costly in the end. 
On the other hand it has to be remembered that the use 
of electrical detectors, leading as it does to fine work, gives 
trouble at times, the source of which may be difficult to 
trace. The Hall Road disaster of July, 1905, on the Lanca- 
shire and Yorkshire R. was in part due to the signalman, 
when he found that he could not get his main line signal 
lever, assuming that the cause for this was " that the electric 
lock had got out of order because it had prevented me from 
pulling the inner home signal off about two months pre- 
viously." Whereas the real reason was that the road was 
set for the middle siding. 

Sykes' Detector. 

One form of electric detector for ensuring that the 
switches of points are in their proper position is one of the 
illustrations (fig. 570) given in connection with the description 

of the signalling of St. Enoch's Station, Glasgow. Two 
other forms used at Victoria (L.B. & S.C.R.) Station are 
illustrated by figs. 593 and 594. The same object is obtained 
in a different way by the use of another detector designed bv 
Mr. Sykes. 

This, fig. 129, is contained in an iron case a attached to 
the rail. To the switches of the points is connected the arm 
b, and to this is attached by the rod c a crank with arms 

d 1 d*. The arm d- has upon it 
springs e l e 1 which form con- 
tact with contact plates / / or 
g g. As seen in the illustra- 
tion contact is made with 
plate / /, and when over, as in 
' dotted lines, contact is made 
with g g. 

On the levers in the locking 
frame of those signals affected 
are locks similar to those 
shown in fig. 574, and when 
the signalman wishes to 
lower a signal he presses a 
button or uses a foot contact, 

which causes a current to flow 
Fig. 129. Sykes Electrical Detecloi ... 

through the plates / / or g g 

as the case may be, which, if 

switches are properly home so that the plates are joined 
together by the springs ' e\ will take the locks out of the 
tappet on the signal lever. Should the switches, or either 
of them, not be properly over, then the current stops at the 
detector and the necessary release of the signal lever is not 

The same result might be attained by the use of the 
switch instruments illustrated by figs. 364-368. 

McKenzie and Holland's Detector. 

This arrangement, fig. 130, provides means for not only 
detecting that the switch points are in their correct positions 
and properly secured in that position by the locking bolts or 
plungers, but that the usual locking bar has been moved. 

Fig. 130. McKenzie and Holland's FJectrical Detector. 



Two locking bolts or plungers 3, 4 are provided for each 
pair of points, one for locking the points in their normal 
position and the other for locking them in their reverse posi- 
tion. These plungers are pivotally connected at one end to a 
sway beam 5, and their other ends are adapted to engage 
their respective notches in the usual split stretcher-bar 2, 
electric contacts 30. 40 being arranged in the plungers 3, 4, 
and co-operating with spring contacts u, 12, connected with 
a battery 18 and electric locking and indicating devices 19, 
iga. A clip 20 fig. 130 is provided with a notch 22 on its inner 
edge, with which a tappet lock 23 is adapted to engage, the 
tappet lock working in a housing and being normally pressed 
outwards, into position to engage the notch or recess 22, by 
means of a spring 25. As shown, the tappet lock 23 is pro- 
vided with a pin or projection 26 about which spring 25 
works. A stud 27 on the tappet lock 23 engages in a slot in 
the lever 28 pivotted at 29, the lever 28 carrying at its upper 
end a metal contact piece 30, which, when the lever 28 is 
operated, makes contact with a pair of springs 31 which are 
connected with the circuit wires in any convenient or known 
manner, the springs 31 being mounted on an insulating sup- 
port 32. 33 is an electro-magnet connected to the circuit 
wires and battery 18 through terminals 34, the armature 35 
of the magnet being pivotted at 36 and provided with a lock : 
ing arm 37 which, in the normal position, when the magnet 
33 is de-encrgiscd, engages a notch 38 in the tappet 39 which 
is pivotted at 40 on the lever A 1 . 41 is a guide roller or sup- 
port for the tappet 39. 42 is an indicator which is operatively 
connected to the short arm 43 of the armature 35 by means 
of rod 44. 45 is the casing in which the parts are mounted, 
the casing being itself mounted on or adjacent to the inter- 
locking frame. 


For working cither of two signals by one lever, electrical, 
instead of mechanical, selectors may be used. 

Sykes' Selector. 

The general arrangements of that designed by Mr. Sykes 
are illustrated by fig. 131 and the details by fig. 132. In the 
case under notice it is where a draw-ahead signal is under a 
right-away signal. The upper arm is worked by the rod a 1 , 
and the lower arm by rod b 1 , and these are coupled by short 
rods a 2 b 2 respectively to a lever c working in a slot d, and 
actuated by the usual balance lever e. Into the short rods 
a 2 b 2 are bolts /' f- which are connected together by the 
spring g. These bolts are so shaped that they can be pressed 
inwards and out of the short rods a 2 b-, should pressure be 
exerted by the raising of the short rods, but they both can- 
not be pressed in, as lying between them is the tumbler h. 
This works on the lever /, and its position is determined by 
whether or not the armature fe is attracted by the magnet I. 
As shown in the illustration the armature is not attracted, 
and consequently, when the lever is worked the short rod b- 
could be raised, as the bolt f 2 is free, whilst bolt / is held by 
the tumbler h. Should the armature be attracted, the tumbler 
h would be tilted, and so hold the bolt f- and free bolt f 1 , so 
that when the lever is worked, short rod b 2 would be held but 
a 2 would be free, and the upper arm could be raised. 

In the case under notice, the 
electro-magnet is controlled by 
the state of the line ahead, which 
determines whether a clear or a 
draw-ahead signal shall be given. 

Other forms of electrical selec- 
tors are those used in connection 
with the Westinghouse, the Low 
Pressure Pneumatic, the Taylor 
and other systems of power inter- 
locking described in Chapters 
XIX., XX. and XXI. 

Fig. 132. Details. Fig. 131. General View. 

Sykes' Electrical Selector. 

The electric selector can be put to the uses shown in fig. 
133. In all the examples there given an electrical connection 
is given from the point c, which will determine whether sig- 
nal a or b will be lowered. 



Fig- '33- 


Use of Electric Selector. 

The Board of Trade object to selectors being used for 
signals that lead to converging movements. They have a 





Fig-. 134. Undesirable Selection. 

fear lest both signals should come off together and lead to a 
collision. Such cases are indicated by fig. 134. 

Fouling, Clearance or Train Protection-Bars. 
Electricity may also be used for ascertaining whether a 
road be fouled before the signal is lowered for a train to pass 
on to that road. 

This is generally done by a bar similar to those described, 
earlier in this chapter, for unlocking purposes (p. 58). There 
is the usual lock on the lever, which is electrically connected 
to a contact plate. Should a vehicle be standing on the bar 
it would be depressed, and so draw the springs away from 
the contact plate so that the current for withdrawing the 
lock from the signal lever is broken, and the signal cannot 
be lowered. 

The Author recommends that " Track-Circuit " be tried 
as an alternative to fouling bars. One advantage is that any 
length of road may be dealt with. With electrical bars any 
length greater than 40 or soft, has to be protected by more 
than one bar, but in " Track-Circuit " the length is imma- 

Take for example the junction of a double line with four 
lines of way shown in fig. 135, which illustrates a junction 
that the Author is acquainted with. It is possible for a train 



Fig. 136. 


to stand with its tail between a and b, or between c and d, 
and under those circumstances no train should proceed from 
X or Y to Z. Were electrical fouling bars used, they would 
have to cover a space of about 150 ft. on each of the roads 
A B. Instead of putting such a length of bars down, all that 
is necessary is to make the line from a to b, and from c to d 
into " Track-Circuits," by putting insulated joints at a b c d, 
bonding the rails in those lengths together, and then by a 
contact maker on the signal lever, sending an electric current 
through the rails which would take the locks out of the signal 
levers. Should any vehicle be standing on the insulated 
portion, the current would be short-circuited, and the locks 
would not be withdrawn. 

In a terminal station like that illustrated in fig. 136, a 
train having to enter road D would have to have the lengths 
a b, c d, e / clear. This would mean innumerable bars, but 
three lengths of " Track-Circuit " is simple. 

Again, in fig. 137, the junctions between the lines A B 
and C D have, owing to the curve on the lines C D, to be 
necessarily of great length, and there is a fear that whilst a 
train was travelling from A to C, the facing points in line 
A, after the train had passed over them, could be reversed, 
and a train come from B whilst the line from A to C was 
foul. A series of bars from a to b is out of the question here, 
but it is very simple to lay a Vength of " Track-Circuit." 

Electric Control of Signals. 

Whilst not germane to the subject of " Lock-and-Block " 
this is probably the most convenient place in which to refer 
to the electrical control of signajs at one box by the signal- 
man at the box in advance. 

Such control is exercised, for instance, where a series of 
"short-sections" follow each other and where it is not 
desirable that the distant signal at the entrance to a series of 
" short-sections " should be lowered unless all the stop sig- 
nals in the sections be also lowered. 

Such an arrangement is in force at Slough on the Great 
Western R., where the signals are placed as shown in fig. 
138. The block sections here are all short ones, and there- 
fore the lever in the Middle Box for working the main line 
distant B is controlled by the lever working the West Box 

Kiddle Bo* 


UTD- i 

-grf" f/eftased tf F C. 

jgr- 137- 

Examples of " Track-Circuit " for Train Protection. 

U W J ' 


Fig-. 138. Electric Control of Signals. 

distant C. The Middle Box branch distant D is similarly 
controlled by the lever working Bath Road distant E. The 
levers working Middle Box distants B D further control East 
Box distant A. As a consequence the latter signal cannot be 
lowered unless there be a clear road indicated by all the stop 
and distant signals in advance being " off." 

Bonding Rails. 
Rails are bonded together by short lengths of iron wire, 



Sykrs" Wire-Bonded Joint. 

small holes being drilled in the rail through which the end of 
the wire is passed and secured there by a channel pin. 

In order to avoid the drilling of the rail, the YV. R. Sykes 
Interlocking Co. have introduced the clamp illustrated by 
fig. 139. All scale and rust is removed from that part of the 
rail where the bond wire is held against it by the clamp. The 
wire a is placed in the groove b and pinched against the rail, 
when the screw c is tightened. 

Pail Drills. 

The drilling of rails for bonding purposes is a tedious 
task when done by hand. It may, however, be done by 

The Wilson drill will make a hole from ^in. to fin. diam. 
by means of the machine illustrated by fig. 140. The driving 
gear is of the sprocket and chain type. On the crank shaft 
are t\vo sprocket wheels. One, fastened rigidly to the shaft 
and revolving with it, drives the feed nut on the drill spindle. 
The other, placed loosely onl the shaft, revolves .with it only 

Fig- 140. 

when engaged by a pawl on the end of the shaft. This 
wheel, when engaged by the pawl, drives the drill spindle- 
so that the drill spindle and the feed nut both revolve in the 
same direction but so geared that the feed nut travels a little 
faster than the drill spindle and so imparts to it an even and 
continuous feed either forward or back, as the crank is 

Another feature is that, by disengaging the pawl, the 
sprocket wheel which drives the drill spindle is made sta- 
tionary while the feed nut revolves, imparting a quick for- 
ward or back movement to the drill spindle. This pawl is 
disengaged when the machine is placed on the rail, and, by a 
few forward turns of the crank, the drill is moved quickly 
forward until it comes in contact with the rail. Then, by 
simply engaging a pawl, the quick feed is cut out and the 
slow automatic feed is brought into action until the hole is 
drilled. The pawl can then be instantly disengaged, thus 

Fig. 143 

cutting out the slow feed, and a few backward turns of the 
crank will retract the drill bit. 

The machine is also shown, fig. 141, thrown back to de- 
tach it from the rail. The lever on the, right is thrown back 
so that the frame can slide and clear the drill point from the 
rail and allow the machine to be removed in a few seconds. 

The Paulus drill fig. 142 is of the same type. It can 
be thrown back clear of the rail, as seen in fig. 143, without 
disturbing the drill. 

Mr. \V. A. Green, of Norfolk Street, Strand the Euro- 



pcan agent for the Buda Foundry Co., who manufacture the 
\\ ilson and Paulus drills has made an improvement in the 
Wilson drill shown in fig. 140, and by means of which a rail, 
provided with a check rail, can be easily drilled without 
removing the check rail. 

Insulated Joints. 

For dividing one section of "Track-Circuit" from 
another the rails of one section have to be isolated from the 
other by insulated joints. 

Weber Insulated Joint. 

In America the joint most used is the \Veber. It is 
manufactured by the Rail Joint Co., and as used in America 
is illustrated by fig. 144. 

Fig. 145 illustrates the Weber joint adapted to the bull- 
headed rail, and as in use upon the L. and South Western R. 

The joint is 34ins. long, and has six Jin. oval necked 
fish-bolts which pass right through the oak fish-plates. 

enter the fibre and prevent them moving on the fibre. The 
insulating bed is of fibre ; it is ^in. thick, and is carried up 
the inside of the chair, but outside the outside oak fish bar. 
The "chair" or "shoe" rests at each end on "raiser" 
plates, which are fastened to the sleeper, and which are for 
the sole purpose of giving the inclination of i in 20 to the 
rail. Between the ends of the rails arc two insulating fibre 
" end posts," each {in. thick, and which rest upon the fibre 
bed under the rail. 

The fish-bolts are insulated at each end ; the fibre bush- 
ings and their protecting metal washers are shown in detail 
on the drawing. Formerly the bolts of these joints were 
insulated where they pass through the web of the rail, but 
that plan was abandoned in favour of the one illustrated, 
because the bushings wore rapidly, and could not be renewed 
without taking the whole joint adrift, whereas in the plan 
shown, the bushings very seldom require renewing, but 
should one fail it can be renewed without disturbing the rest 
of the joint. 

Sykcs' Insulated Joints. 

The W. R. Sykes Interlocking Signal Co. have an insu- 
lated joint, which is illustrated by fig. 146. 

!"'& '44- AVtber Insulated Rail Joint for Flange Rails. 

The rail rests directly upon two bed plates iGins. x 
4jins. x jin., which do not touch in the middle, and which 
protect the insulating bed on which they lie from chafing, 
due to the movement of the rail. They have snags which 

Fig. 146. SyUcs' Insulaled Joinl. 

Each fishplate is made in two parts, one end of each fish- 

I \ ^r"^ y ( i i 


Elevation Outside of Shoe 

16 ~ 

This hole on opposite side in 
right hand raiser plate 


This hole on opposite side m 
right hand raiser plate 

k 4% 


Detail of left hand raiser plate 
Fig. 145. Weber Insulated Rail Joinl for Bull headed Rails. 


plate being turned as shown. Insulating material is placed 
between the two rail ends and between the turned ends, and 
two bolts are passed through fishplate and rail. The two 
ends are also fastened together by two bolts, the latter being 
insulated from the plates by fibre washers. 

These joints are used on the Gt. Northern R., the Great 
Central R., the whole of the M. District R., and on other 

Atlas Insulated Joint. 

Fig. 147 illustrates the "Atlas " insulated joint for the 
loolb. Am.Soc.C.E. standard rail. P'our fishplates, which 
are secured to the ends of the rails by two bolts through each 
rail-cnd, are used. Two bolts under the rail also couple the 
fishplates together. Insulating material is placed between 
the rail ends, between the fishplates and rail and under the 

Section A B. 

Fig 147. Atlas Insulated Rail Joint. 

rail. The bolts are provided with fibre bushings, which have 
a steel spring round them to assist in protecting the insula- 
tion from wear. 

Kohn Insulated Joint. 

Mr. K. Kohn, of 45, West ngth Street, New York, has 
designed the joint illustrated by fig. 148 with a view to 

Fig. 148. Kohn's Insulated Rail Joint. 

eliminate any chance of the insulating material being 

The insulating bars a are made of hicque a very hard 
wood grown in Cuba and two of these are placed in two 
fish plates b and the rail then built up as shown. The slot c 
in the centre of the insulating bar a is to hold the piece of 
the insulating material that is put between the rail ends. 



Iris sometimes, when constructing- a railway, impracticable 
to avoid its crossing navigable waters at a level too low to 
allow ships to pass, and in such cases an opening bridge has 
to provided in the railway. Such openings have to be very 
carefully protected and interlocked with the signals so as to 
secure a safe passage for the trains when the bridge is closed 
and also to clearly indicate to approaching trains when the 
bridge is open for ships to pass through. 

It is not only essential to ensure that the bridge is right 
for railway traffic before a train is accepted, but that the 
rails are in alignment and that the moving and fixed struc- 
tures are level with each other. It was the absence of the 
latter that led to the Atlantic City disaster on Oct. 28th, 
1906, when 56 lives were lost owing to a train being derailed. 

The signals at the boxes on each side of the opening 
bridge should be controlled from the bridge so that they can- 
not be lowered after permission has been given to open the 
bridge. Block communication broken down so 
that no trains may be signalled, but telephonic communica- 
tion must remain in use. 

Notwithstanding the thick network of lines and the large 
number of navigable waters there are not more than 40 
opening bridges on the railways of the United Kingdom, 
and all these are protected by fixed and block signals on the 
plans described below. In America opening bridges are pro- 
portionately much more common, and they are not as a rule 
so carefully protected, and consequently one still reads of 
trains " falling through an open-draw " as at Norfolk, Va. , 
in the summer of 1905. 

Where the " open-draw " is protected in America it is 
generally done by providing stop signals at some distance 
from the bridge, also distant signals at the regulation dis- 
tance from the stop signals. As the block system, as under- 
stood in our country, is not generally used in America (where 
automatic signals are used instead) it is not possible to break 
down the block, so reliance has to be placed entirely on the 
signals which are interlocked with the opening bridge, and 
also with derails or facing safety points and the tower- 
man (signalman) is warned of the approach of trains by indi- 
cators working in connection with the automatic signals. 

In this country signal-boxes are generally provided at the 
bridges, together with fixed signals, and before the signal- 
man at the bridge box can withdraw the bolt controlling the 

draw, permission has to be obtained from the signal boxes 
on both sides of the bridge, and the granting of this per- 
mission by the men in the distant boxes breaks down the 
block instruments, so that no train can be signalled towards 
the bridge, and the fixed signals leading in that direction are 
locked in the danger position. 

Telephones are always provided, which are free for com- 
munication at any time. 

Amongst the swing bridges in the United Kingdom the 
signalling of which are examples of what is done are the 
following : 

Neath River Bridge. 

One of the most complete arrangements of such inter- 
locking are those at the Neath Bridge on the Rhondda and 
Swansea Bay R., and which is a double line. The work was 
carried out by McKenzie & Holland, Ltd., and the Author 
is indebted to Mr. W. Sutcliffe Marsh, the engineer, for the 
particulars now given. 

Fig. 150 is a diagram of the lines and signals at the 
River box and at the two boxes on either side Dynevor 
Junction and Neath Junction and figs. 151, 152 and 153 are 
sketches of the electrical arrangements at each of the boxes 
named. These are given in complete detail and explain 

Barnstaple Bridge. 

In fig. 154 is given an example of the interlocking of a 
swing bridge oa a single line. This is at Barnstaple on the 
L. and South Western R., where the bridge is controlled by 
Sykes' lock and block in conjunction with Tyer's electrical 

There are three boxes concerned : Barnstaple Town, 
where the bridge is situate, and Barnstaple Junction and 

The section of single line is from the Junction to Pot- 
tington, and these are the two tablet stations. The Town 
box is not a tablet station, consequently, although the bridge 
is between the Town Station and Pottington, the arrange- 
ments extend to the Junction. 

The bridge is controlled by No. 4 lever in the Town box. 
When an up train (i.e., one from Pottington to the Junction) 
has to pass over the bridge, the Town man pulls No. 4 lever, 
and then lifts up the side lever of the plunger of his block 


To accept a train from Ncath Junction, No. 17 must be first pulled 
over, then press plunger (which releases Neatli's starting signal and 
locks 17 in the over position, and is released by the accepted train pas- 
sing over the treadle A. 17 is now returned to tlte normal position, and 
must again be pulled before accepting another train. 

To 'send a train to Neath Junction, No. 6 starting signal lever must 
be released from Neath Junction. 

The River Bridge is released by pulling over No. S point*, lifting 
handle under shelf (which lacks 2, 3, 5, (> and 17, and 8 in the over 
position), and then plunge on instrument. 

Fig. 151. Electric Locking, Dynevor Junction. 




The key lever is released by electric instruments. Key lever when 
pulled over releases bearing block lever, and lifting lever and back locks 
electric instruments. Lifting lever or bearing block lever pulled over, 
back locks key lever and withdraws the blocks, breaking contacts ABC 
and D, which lets the electric lock E fall into the notch F, and allows 
arm of indicator G to fall from the In position, and the arm indicator 
H to jail from the Out position. 

The Bridge on being swung breaks contact O, the object of which is 
to still block the key lever " over," should the bearing blocks be restored 
to the In position when the bridge is open. 

Fig-. 152. Bearing Block Indicators and Lock, Neath River Bridge. 

To accept a train from Dynevo 

j<i must 



then press plunger (which releases Dynevor's starting signal), this locks 
39 in the oz'er position, and is replaced from this by the accepted train pass- 
ing over treadle B. 39 is twiv to /he normal position, and mnst again be 
pulled before accepting another train. 

To send a train to Dvnevor, No. 42 starting signal lever must be 
released from Oynevor Junction. 

The River Bridge is released by pulling over 24, then lifting handle 
under shelf (which locks 39 and 42, and 24 in the over position), and 
then plunging on instrument. 

Fig. 153. Kiel-trie Locking, Neath Junction. 

instrument, whereby No. 4 lever becomes locked, and presses 
in the plunder. The tablet instruments at the Junction and 
Pottington are now unlocked, so that a tablet can be issued 
in the usual way for a train to leave I'ottinyton for the Junc- 
tion. No. 4 remains locked until the train has passed over 
the electrical treadle A on the Junction side of the bridge, 
which lifts the lock out of No. 4 lever. 

For a down train No. 7 starting signal has to be lowered. 
This lever is locked by the side lever of the plunger of the 
block instrument, and the same movement is gone through 


Fig-. 154. Barnstaple Bridge, London and South Western Railway. 

Was. in Circle are worked from Barnstaple Town. 

\ BOX 



as already described, so that when the side lever is lifted No. 
4 is locked as before, the tablet instruments are freed and No. 
7 is unlocked so far as the Town Station is concerned. No. 
17 lever at Pottington is then pulled over and the signalman 
there plunges his block instrument to the Town Station, 
which frees No. 7 signal there. The bridge bolt lever No. 4 
at the Town and No. 17 signal at Pottington remain locked 
until the train has passed over treadle B at the Pottington 
side of the bridge. 

The levers working No. 8 starting signal at the Junction, 
and No. 5 starting signal at Pottington, cannot be pulled 
over until a tablet has been taken out of the instrument, so 
that a signal cannot be lowered for a train to approach the 
bridge unless a tablet is out. This guarantees that the line 
is clear before the signals are lowered. 

Security is thus obtained by No. '4 at the Town locking 
the bridge. This frees the side lever of the plunging instru- 
ments in both the Junction and Pottington boxes. The 
lifting up of the side lever unlocks the tablet instrument, and 
a tablet in its turn unlocks the starting signals for proceed- 
ing towards the bridge. 

Shannon Bridge, Drumsna. 

This bridge is situate between Drumsna and Dromod on 
the Midland Great Western R., Ireland. This portion of the 
line is single and is controlled by the electric train staff. 

The bridge is protected by a signal box with four levers 
(for working up and down home and distant signals), and a 
fifth lever for unlocking the bridge, which is a bascule lifting 
bridge. The arrangements are shown in fig. 155. 

The bridge lock lever No. 3 is unlocked by a special 
switch lever, which is held by an electrical lock on the staff 
instrument wire. When it is necessary to open the bridge, 
the man there exchanges a special bell code with Drumsna 
and Dromod boxes, and, if the line be clear, the men there 
hold down the tapper keys of their staff instruments, and 
the bridge-man holds down the tapper key connected with 
the switch lever. When the indicator at the bridge shows 
" unlocked," he signals to his neighbours that he is free, 


SKIoitBOU * Li I * / . |l7g? 

and the staff wire is then broken down so that no train can 
pass on to the section. 

When the operation is completed and the bridge restored, 
the man there puts back No. 3 lever, and on exchanging bell 
signals, No. 3 lever is again electrically locked, and the staff 
wire re-opened between Drumsna and Dromod. 

Severn Bridge. 

When the Severn and Wye R. was taken over by the 
Great Western and Midland companies the signalling and 
interlocking of the opening portion of the Severn Bridge 
where it crosses the Berkeley ship canal were overhauled by 
the Great Western Co. 

This is a swing bridge situate between Sharpness North 
and Severn Bridge Station boxes. It is on a single line con- 
trolled by the electrical tablet. A signal box is provided on 
t'ic opening portion. The signal boxes, i.e., Sharpness 
North and Severn Bridge Station, electrically control each 
other's starting signals for leading towards the bridge, and 
the circuits for the same, also for the tablet, pass through a 
switch ir. the Bridge box. 

A bolt, worked by a lever in the locking frame, and which 
is therefore on the moving part of the bridge, is driven into 
a hole in the ironwork on the masonry pier. This hole fixes 
the exact position the bridge must be in to secure alignment 
between the rails on the fixed structure and on the opening 
portion. Consequently, unless these be in accurate line the 
bolt cannot be shot. 

The opening portion is held level by wedges, which have 
to be withdrawn so that the bridge may be lowered before 
it is turned. These wedges are withdrawn by a screw 
operated by a friction wheel. There are three friction wheels. 
One in the centre is coupled to the engine, and this engages 
with either of the other two. One of these latter is for the 
screw mentioned above, and the other is for swinging the 
bridge. The connection to the middle wheel is bolted from 
the signal box and this, of course, controls both the lifting 
and turning. 


M attached, i? eerdreput of Bridge 


Misenpyers for vires 

\ 1 1 ^ 1 

L, ^*i for con,lrollmq Mndqe f" 1 " 



/?(70 JOINTED-* 

/*r(7^i, Mu-llinyar 

W/5 frtfl iA^J ^ 

7S J^o 

5 K) IS 20 25 30 35 

40 45" 




F'&- '55- Shannon Bridge, Drumsna, Midland Great Western R. 



The switch through which the electrical controls on the 
starting signals and the tablet circuit pass is operated by 
another lever in the locking frame. When the switch is 
open and the starting signals free and tablets may be with- 
drawn, the lever in the locking frame is over so that it must 
be reversed and put to normal in order to close the circuits. 
The bolt-lock lever is interlocked with the switch lever so 
that before the latter can be pulled after the bridge has been 
used for traffic, the bolt-lock lever has to be back in the 

When the bolt-lock lever is put back a lock falls auto- 
matically into the tappet. When the bridge may be opened, 
the switch lever must first be put back, which locks the 
starting signals and breaks down the tablet circuit, and then 
the signalmen at each end must press down the tapper keys, 
and the plungers on the tablet instruments, which sends four 
electrical circuits into the Bridge box, and operate four relays 
that cause a local circuit to take the automatic lock out of the 
bolt-lock lever. 

Breydon Viaduct. 

In July, 1903, the Midland and Great Northern Joint 
Committee opened their connecting line between Great Yar- 
mouth and the Mid. G.N. and Gt. Eastern Joint line to 

This crosses the estuary of the Yare, generally known 
as Breydon Water, by a steel viaduct of five spans, of which 
one is a swing span. 

With the exception of the portion over the viaduct, the 
line is double. The bridge is single and is controlled by the 

road traffic and the lower for railway over which the Great 
Western run. 

It is a double line of railway and there is a signal box on 
each side of the bridge Ashton Swing Bridge North and 
Ashton Swing Bridge South and a locking frame is fixed 
in the Bridge house. The signalling at the swing bridge 
was carried out by Saxby and Farmer, Ltd. for the Docks 
Committee, and, of course, the Great Western Co. did their 
own work. 

Safety points are provided in each line on both sides of 
the bridge (the lines are not used for passenger traffic). 

In fig. 156 is a diagram of the levers, and their purpose, in 
each box. 

The bridge is electrically bolted from each of the Great 
Western boxes by a lever in the locking frame which is inter- 
locked with conflicting levers. 

The levers working the home signals leading on to and 
off the bridge, when they have been used, may be put back, 
but not fully. They cannot be placed completely to normal 
until the whole of the train has passed off the bridge and 
gone over an electrical contact. This contact is not made 
by the first wheel of the train, as the wheel first depresses a 
weighted bar which switches out the contact. When the last 
wheel has passed off the bar, the latter rises so that the con- 
tact is switched in, and when the last wheel depresses the 
contact a circuit is completed which allows the lever to be put 

fully back. 

Fig 157 is a wiring diagram of the Great Western boxes, 
from which it will be seen that the bolt-lock levers cannot be 

ving been pulled over cannot be completely 
until Train has passed m'cr 
but can be put back sufficiently to put 

Levers A'tw. 18 &* 19 and 3 or 3 4 ka 
returned to Normal Position 
'* A " and " B" respectively ^ 
Signal at Danger. 

ers Nos. 12 and f. after 
having been pulled over, cannot he com- 
Hedge plctely returned to their Normal t osition 
until the train has passed over treadles 
'A" and '/>''' respectively, but can be 
put tack sufficiently to put the Signal 
to danger. 

Bolt releasing wedge and lever controlling power. 

electrical tablet operated by signal boxes on each side of the 
bridge. There are levers in the operating house at the bridge, 
and the bridge is also controlled by levers in the adjacent 
signal boxes. These latter levers are normally " over " in the 
frame and the bridge locked. When reversed the signals 
are interlocked and the tablet circuit broken. The move- 
ments at the Bridge box are led by a controlling lever which 
releases the lifting and turning levers, and these, naturally, 
" back-lock " the controlling lever. 

Ashton Swing Bridge. 

This bridge was erected by the Docks Committee of the 
Bristol Corporation. It has two levels, the upper one for 




South Bo.res can only 
be inserted when the 
briilge is set J'or the 
railivav and locked. 

Fig. 156. Ashton Swing Bridge. 



r,d } , , 

drco'or- IGr*** Irfttfi a- Ht.:1t ywnef 
j5Sjf I BOlTfo (tjc.--f.rl 

fz? /I a/=< ,- /wy,,,/ 
p*tf ^(wsoiT-fo it,?,, i, i 

Fig. 157. Wiring Diagram, Ashton Swing Bridge. 

restored until the wedges and bolts operated by the Bridge 
box are fully home. 

Barrow and Suir Bridges. 

The new line from \Yaterford to Rosslare, and the new 
connection over the Suir that brings the Mallow and Cork 
line of the Great Southern and Western R. into the North 
station at Waterford, necessitated the erection of two long 
bridges only exceeded in length, in the United Kingdom, 
by those over the Forth and the Tay. 

Both bridges have opening spans. The one over the 
Barrow (about 5 miles east of Waterford) is of the usual 
swing type, while that over the Suir (immediately west of 
Waterford) is a Scherzer rolling-lift bridge. The signalling 
and interlocking to control the working of these bridges was 
carried out by Saxby and Farmer, Ltd. 

Fig. 158 is a diagram of the cabin at the Barrow Swing 
Bridge. The six levers on the right-hand side of the locking 
frame are for bridge purposes and control the operations 
in the machine room. Home signals are provided in each 

I 1 1 1 Starting Signal 

direction, that from Waterford being repeated by an. outer 
home signal owing to a tunnel intervening. The whole of 
the new line is worked by Tyer's tablet instruments, and the 
signalling has been so arranged that no tablet can be with- 
drawn if the bridge is open or is being opened, and the 
bridge cannot be opened if a tablet be out. 

Above the locking-frame are the block instruments which 
work on the east side to Campile and on the west to Abbey 
Junction, Waterford. Lever C is the master lever, and this 
is controlled from both Campile and Abbey Junction boxes. 
The starting signals at those boxes for the direction leading 
to the bridge are electrically controlled from the Bridge 
cabin. Then the master lever is interlocked with the fixed 
signals at the bridge, so that, so far as the fixed signals 
are concerned, the operations are efficiently protected. But 
more is done. The master lever, when pulled, cuts off by 
means of an electric contact maker the tablet wire current, 
so that no tablets can be withdrawn, but when a tablet is 
out the master lever in turn is locked and cannot be pulled. 

/fbbey Junction 


Indicator and RelL Indicator and Bell. 


le Cabin. Abbey Junction Cabin. 

rr TJ 

^ll < 

^5 " f bridge c<ibut. campile. \ 
f ( f*'l ntnr worked Tndiwtor .$>>/* \ 



Gong rung fro 

! Indicator worl 
show when 

71 ^fe&y Junction. bridge cabin. \ . -^V 
fndirntor v>orlf<l . 


V c,- 5 S 

~TT from bridge cabin by campile to \ 
i tn show when thr show when the '- 
1 starting signal is key lever ** C" 

ed by Abbey Junction to from bridge cabin \ -j 
the key lecer " C" it to show when they "=j 
starting signal \ 

(D- =|1 

; 5 

i -c o 

*5 3" 

-O^= _ .treil. isfrejd. 
( Plunger to ring the Plunger and han~\ 
-. bell at bridge die for working \ 
( cabin. the indicator and \ 




( Plunger and 

-L indicator an 
V starting sign 

bundle for working the p } emT '"/"'^ ,h f { =5" 
i bell for controlling the Pl'ffer to nngth?) 


,1 at Abbey Junction. ^ "' '"^ / 

ling the starting I 
signal at cam- \ 
pile. J 




Apparatus in 
Lever. " A " for signal A. 
, ** B" for signals B B. 


, *'<?'* i* iA A-tfy (or X:in^) /;r. 


) n 




, la locking bolt lever. 


, 2. L\fti*g lever. 


, 5. Blocking 


, ^. Turning 


t 5. Latching ,, 

A I! C 

la 1 2 3 i 5 

Fig. 158. Diagram of Signals and Interlocking, Swing-Bridge Cabin, Barrow Viaduct. 






L iff ing Span I Ufa Bridge Cabin 

H 4O2 Yds. 

Gracedimu Junction 

Indicator and Jtell. 
Grace/Hen Junction Cabin. 

Junction Home 


CaUn C. bl*rford.N. 


( Jiett rung from 
) bridge cabin. 

Indicator and Sell. 
Cabin C, Waterford. 

1 rung from Grace- \ 
(tifu Jinn-lion. \ 

Q* [Indicator worked Indicator worked by \ 

I from bridge cabin Gracsdieu Junction ' 
to show when the to show when key f 

starting signal lever is freed. 
\ lever is freed. Plunger and handle for \ 

working the. indicator f 
and bell and far con- '- 
trolling storting sig~( 
nal at Gracedieu. ) 

i Plunger to ring 
the bell at bridge 
\ cabin. 

6 Levers and Winch in Cabin. 
Jwo levers for railway signals. One lever 
for locking bolts and pawls at nose end of 
bridge. One lever for locking bolts and 
pawls at rear end of bridge. Ont lever for 
sM,ng bars for keying rails level and in line. 
One lever as key (or king] lever. 

Winch for 
river ball. 

hoisting )| 


j Gong rung from Tiell rung from ) 

( Ind'rator worked by 
Waterford to show 


key lever is 

Indicator woi'ktd\ *(7) 

from bridge cabin I 
to show when the L - 
junction signal I 

lever is freed. 
IF _ . ' _ 


Plunger and handle. 

for working the in- Plunger to ring , 
dicator find bell and the bell at bridge \ 
for controlling June- cabin, 
tion signal at Water- 



Fig. 159. Signalling and Interlocking for Lifting Bridge, Suir Viaduct. 

A ball signal is provided which is raised when the bridge is 
about to be opened and lowered when it is about to be closed, 
and the winch for working the ball signal is interlocked so 
that it cannot be raised until all the operations have been 
gone through that are necessary. The result is that river 
craft cannot be misled. 

When the bridge requires to be turned the bridgeman 
pulls lever C (the master lever) half-way over. This he can- 
not do if there be a train in the section or a tablet out for 
one. Having got the lever half-way over he asks Campile 
and Abbey Junction for permission to turn the bridge. The 
granting of this permission cuts out the tablet line and frees 
lever C so that it can be pulled fully over. An indicator is 
provided in the Bridge cabin to show when this permission 
is received. The bolting levers, i and la, which stand nor- 
mally over, are interlocked with lever C, so that when the 
latter is pulled the former can be put to normal, and these 
in turn free lever 2, the lifting lever, so that the ends of the 
bridge are lifted ready lor turning. The performance of this 
is secured by an electrical lock on the next lever to be moved 
No. 3, the blocking lever which lock is operated by the 
lifting mechanism, so that until the ends have been lifted the 
blocks cannot be withdrawn. The complete withdrawal of 
the blocks enables No. 2 lever to be reversed, allowing the 
bridge to again fall, and this frees No. 4 lever so that the 
bridge is furned and is then latched in the open position by 
No. 5 lever. 

To- elose the bridge the order of the movements is re- 
versed, but lever C cannot be restored to normal until an 
electrical contact has been made showing that the bolts 
worked by levers i and la are fully home. 

The arrangements at the Suir Bridge are practically the 
same, except that they have to be modified for controlling 

a rolling-lift bridge operated by hand instead of by power, 
and as the section of line on which the bridge is situated is 
worked by the electrical train-staff, the arrangements have 
had to be modified in that respect also. The diagram, fig. 
159, illustrates the arrangements. 

American Systems. 

To the Union Switch and Signal Co. the Author is in- 
debted for the diagram, fig. IMO, showing the signalling and 
interlocking of a drawbridge that was carried out by them. 

In studying the diagram it must be remembered that 
American trains travel on the right hand road and signal 
arms point the opposite way to British practice. 

The signal-box or " tower " is placed on the bridge and 
contains 8 levers, i of which is spare. Five hundred feet 
from the bridge are placed in each road facing safety points 
or derails, each worked by a separate lexer and provided 
with facing point locks. These points are interlocked with 
the bridge lock lever so that when the bridge is to be turned, 
the derails must be opened. Stop signals are fixed at the 
derails. 'Track-Circuit" is provided throughout, as the 
bridge is on a line control'ed by automatic signals. The home 
signals are worked from the tower and are controlled by the 
" Track-Circuit." Distant signals are provided on the auto- 
matic stop signals in the rear on each side. These signals arc 
semi-automatic and follow the bridge home and automatic 
stop signals without being worked from the tower. The 
bridge lock lever is electrically controlled by the lowering 
of the home signals, a lock being shot into the lever, which 
is only withdrawn when the train has passed over the end of 
the "Track-Circuit" extending from the stop signal to the far 
side of the bridge. Should the lock be shot in error it may 
be released by the signalman using a slow-acting hand 

1 C ' ' 

^ 500 1 *\*50'+\ 

JL_i ^SU 

r- 50 ~Y" 


NoHh Bound 

! \ 


Draw Bridge 



i- Automata 


~M<>chcinicol uith Electric SM 

D> Machine 

Pig. 160. Draw-Bridge Protection, Union Switch and Signal Co. 

Avroma tic 



Owing to the line being protected by automatic signals 
there are no block instruments provided, electrical indicators 
being fixed instead, which show when trains are ap- 

Chicago River, U.S.A. 

Cn referring to fig. 522, connected with the description 
of the Taylor system of working points and signals by elec- 
tricity, it will be seen that at Sixteenth and Clarke Streets, 
Chicago, there is a draw-bridge over Chicago River. This 
is controlled by No. 124 lever, which, when pulled, raises 
the bolt B, fig. 161, taking a lock out of A, so freeing the 
throttle of the bridge engine. The wiring for No. 124 lever 
is given in the illustration, so that it may be seen how the 
electrical connections to those signals leading on to the bridge 
are broken down. 

Charles River, Boston, U.S.A. 

The controlling of this bridge by the Union Switch and 
Signal Co. is described along with the signalling of the 
Boston Elevated Road. See fig. 313, page 167. 

Harlem River, N.Y.C.R.R. 

\\"ith the experience gained by the Atlantic City disaster 
on the Pennsylvania R.R., already referred to, the New York 
Central, when re-signalling their line in connection with the 
electrification outside New York, paid particular attention to 
the alignment of the draw bridge across the Harlem River. 

rln. Submarine Cable 

Fig. 161. Chicago River. 

Circuit controllers have been placed on the end of each mitre 
lift rail to ensure that these rails are set " home " before the 
signals governing movements across the bridge can be 

Fig. 4023. Signal Box at Thompson, Pa., Pennsylvania Railroad (xc-c pp. 225-226). 



THE safe working of single lines has always been a prob- 
lem. In the early days of British railways trains were 
worked to a programme as to where each was to meet and 
cross others from the opposite direction and which, before 
the telegraph was brought into use, had to be worked to 
with cast-iron regularity. But the introduction of the tele- 
graph allowed instructions to be modified when trains were 
running late or otherwise out of order, and naturally led to a 
wider adoption of these crossing-orders, which were regarded 
by many railway officials as a means of safety, which, of 
course, they were so long as they were properly worked to, 
but there were many accidents, possibly the most serious 
being that at Norwich in 1874, when, owing to a misunder- 
standing as to a telegraphic message, two trains from oppo- 
site directions were allowed to meet, causing the death of 
25 persons. 

Safety in the early days of railways was also secured 
by having a pilotman, who accompanied all trains over the 
single line, and if there were two trains waiting to proceed 
in the same direction he despatched the first and followed 
with the second, or with the last if there were more than two. 
This led to tickets being carried by the pilotman in a pouch, 
which he issued to the trains as an authority to proceed 
without the pilotman. 

The wooden train-staff was subsequently introduced, 
which was a token instead of the personal presence of the 
pilotman. There was one train-staff for each section and 
no train was allowed to proceed into the section unless it 
had the train-staff for that section. 

This system was first adopted on the L. and North 
Western R. near Leamington. Tickets for use with the 
wooden train-staff, kept in a box that could only be unlocked 
by the train-staff, followed, and provided for those cases 
when there were more than one train waiting to go in the 
same direction, or when one or more trains were expected and 
which would require the staff before it could come back from 
the other end. 

The staffs were subsequently made of iron or brass. 
They were differently shaped for each section, so that a 
round staff would not unlock the ticket-box of a square staff 
and vice-versa. The boxes were also so constructed that 
when they had been opened by the train-staff they must be 
closed again before the staff could be withdrawn as the staff 
was locked in as long as the lid was not shut down. 

But even with the addition of tickets the method was un- 
wieldy and caused delays. A train would arrive at the en- 
trance to a section when the train-staff was at the other end 
and it would have to wait until the staff was brought by a 
train from the opposite direction, or was conveyed by hand. 

Consequently a great boon was conferred on English 
railways when, in 1878, Mr. Edward Tyer introduced his 
electrical tablet system on the Caledonian R., as by its use 
trains could be admitted from either end of a section as long 
as there was no train in the section and after such train had 
been properly accepted. 

It removed all the risks of working trains by a ticket only, 
as there was always a risk that more than one ticket might 
be withdrawn from the box and of it (or them) being subse- 
quently handed by mistake to a driver whilst the train-staff 
was at the other end of the section. This was, of course, 
provided against by the positive instruction to the driver to 
see the train-staff when he accepted the ticket, so as to be 
sure he left the train-staff behind when he went away with 
his train, but still there was the chance of an error. 

It removed the delays hitherto caused by having to wait 
when the staff was at the other end of the section, and com- 
pelled the signalman at the far end of the section to concur 
in the sending of the train. 

The instruments employed were also block instruments 
as well as single line safety appliances, and their adoption 
met the Requirements of the Board of Trade as to New 
Lines and the Railway Regulation Act of 1889, as to pas- 
senger lines being equipped with " the requisite apparatus 
" for providing, by means of the block telegraph system, an 
" adequate interval of space between following trains." 

The tablet system also allowed signals to be dispensed 
with at those stations that were not tablet exchange stations 
unless there were road level-crossings. Siding connec- 
tions could also be controlled by the tablet, which thereby 
avoided the provision of signal-boxes and signals, as the 
siding points were worked by a small ground frame which 
was unlocked by, and interlocked with, the tablet. 

The electrical train-staff followed, and now either the 
electrical tablet or the electrical train-staff is used by every 
railway company in Great Britain, only the smaller and less 
important branch lines being operated by the old train-staff. 

Crossing Orders were abolished in Great Britain some 
years ago. They were last used on the Highland R., on 
which they were abandoned in favour of the electrical tablet. 



On this railway there are long lengths of single line between 
the signal-boxes ; the gradients are exceedingly heavy, and 
the weather, at times, very bad ; numerous specials for fish, 
horse and carriage traffic must be promptly run ; trains from 
the south or north may run late and are often in duplicate 
or triplicate, and some of the mail and express goods trains 
require to run long distances without stopping. With these 
peculiar features crossing orders were not able to cope satis- 
factorily. But the principal reason for the change was that 
the responsibility for keeping all the trains moving, prevent- 
ing delays, seeing that the orders were sent and properly 
answered and, above all, the constant fear of making some 
mistake which might lead to disaster proved too great a 
strain for the nerves of one man. 

In an address given before the Conference of Telegraph 
Superintendents of America at Chattanooga in June, 1905, 
the Author likened the Highland R. to an American road on 
account of its mountains, long distances, irregular traffic, 
climatic conditions, etc. Those readers who are acquainted 
with the peculiarities of both countries will probably agree 
with the Author's comparisons. 

The difficulties as to single lines in Great Britain have 
thus been solved, and the freedom and expedition shown in 
the present day workings are such that the conversion of a 
single into a double line is now a rare event. 

What has been achieved in England has also been at- 
tained in India, Africa, South America, Japan, Australia, 
New Zealand and other British Colonies. 

But in North America the problem of operating single 
lines with safety and speed, coupled with economy, cannot 
yet be regarded as settled. There the train-despatcher and 
train-orders continue to be the principal means of operation, 
although hardly a week passes without a serious " head-on " 
collision being reported because of some error of omission or 
commission in the issue or carrying out of orders. These are 
safeguarded in every possible way. A schedule is laid down 
as to what the normal working is to be, and if for any reason 
the train-despatcher has to modify the instructions con- 
tained in the . schedule his orders are issued over the 
train-despatcher's wire to the trainmen of both' or all 
trains concerned. These, after being noted by the conductor, 
are repeated back, and it is the conductor's duty to read them 
over to the engine-man, fireman and brakesman, or the mes- 
sage may have to be noted by the engine-man as well as by 
the conductor before it is repeated. But yet orders are some- 
times undelivered, as the operator fails to put up his train- 
order signal, and so the train runs through. Or they may 
be misread, misunderstood or forgotten. Mistakes may 
arise because a train at a crossing place is not recognised, or 
a misunderstanding may arise from one or both opposing 
trains running in more than one portion, or the headlights 
(" markers ") not being properly seen or exhibited. 

Other weaknesses of the train-order system for singli 
lines might be mentioned, but American railroad men arc 
aware of its faults, and for others sufficient has been said. 

American railroads must not, however, be too quickly 
condemned. Their conditions are altogether different to those 

to be found in Great Britain or anywhere else. The greater 
part of their enormous mileage, and especially in the west, 
was constructed hurriedly, not to meet the demands of traffic 
but to create it. There being no Board of Trade nor other 
similar body to issue and enforce Requirements as to New 
Lines, the construction has been rushed through sparsely 
populated districts and often over mountains and the prairies 
where, even now, to employ a signalman would be to 
ostracise him and practically send him to exile. In the 
original days of these western lines the only means of ' ' secur- 
ing an adequate interval of space between following 1 
trains" was watching "the other fellow's smoke." It 
should be remembered, too, that there are times when these 
single lines in the west are scarcely able to carry the traffic 
when worked even under the ultra-permissive (or most free 
and easy) conditions. Such is the case when stock (cattle) 
are being sent off the prairies into winter quarters or vice 
versa, and when the grain is being rushed to the markets. 
It is, therefore, natural that the officers of such railroads 
should look askance at any method that would appear to 
hamper them or tie their hands in any way, even although 
such is to prevent their freedom from leading them into 
trouble. There is also the objection, characteristic of every 
American, to being put under any restraint whatever. He 
hates to have his liberty curbed, or to be unable to act in 
accordance with what he considers is the right course, and 
therefore any machine for controlling his movements is 
Anathema to him, unless he can " monkey " with it and 
"beat the machine," when his objections are modified and 
he will, whilst having an inward contempt for it, accept it 
simply because he has mastered it. 

The Author puts forward these suggestions, not in a 
critical spirit, far from it, but as his own opinion as to some 
of the real reasons why methods that have been failures in 
other countries, and lead to frequent disasters in America, 
should still be allowed to continue there. That some officials 
are determined that something must be done is proved by 
the fact that the subject has been seriously considered on 
the great system of railways known as the Harriman lines, 
and which include the Union Pacific, the Southern Pacific 
and other important railways in the western States. The 
Harriman combination controls nearly 15,000 miles, of which 
only a few hundred miles are double lines. The Director for 
Maintenance and Operation, Mr. J. Kruttschnitt, informed 
the Author that the programme for 1907 was the equipment 
of 2,745 m 'l es of single line by the Union Switch and Signal 
Co. 's automatic signals, and then their record would be : 

Double line controlled by the block system 150 miles. 
Single line controlled by- 
Hall automatic signals, normal danger 177 ,, 
Union Switch Co. 's automatic signals, 

normal clear ... ... ... 4,264 ,, 

Electric train-staff ... ... ... 109 ,, 

Total single line protected 4,700 miles 

Another company, the Cincinnati, New Orleans and Texas. 
Pacific, or as it is generally called, the Queen and Crescent 
Route, has equipped the whole of their single line from 



Cincinnati to Chattanooga 336 miles with automatic sig- 
nals. It has been stated * that : 

" It is not an exaggeration to say that, for safety of method com 
bined with simplicity of operation, the 336 miles from Cincinnati to 
Chattanooga is not equalled or even approached on any other railway 
in the United States, and probably not in the world. It is an object 
lesson for those who wish to know, as well as those who ought to know, 
what can and should be accomplished on the railways of the United 
States towards securing the safely of their trains." 

The Author happened to know that the writer of the 
article quoted was qualified to speak with authority, and 
therefore took an opportunity of going over the Queen and 
Crescent Route in June, 1905, and desires to here place on 
record his very high opinion, not only of the signalling, but 
of the equipment generally. It is not easy to put a monetary 
value on the advantages secured by such expenditure on sig- 
nalling, but it is perhaps enough to say that whereas the Queen 
and Crescent R. used, according to legends, to be credited 
with three collisions a day, it is now an established fact that 
they occur only at long intervals. Further, the carrying 
capacity of the line has been increased 30 per cent. 

One difficulty in the way of the adoption of such methods 
as the electrical tablet or electrical train-staff is that operators 
would be required to attend to them. This was the problem 
that was put to the Author by the General Manager of an 
American railroad. He had two schemes before him for sig- 

date. There are thus advantages in having the human 
agent instead of a machine, be the latter ever so perfect. It 
is impossible in electrical methods for opposing trains, as 
long as the rules are obeyed, to be between two crossing 
places at the same time ; there is no waiting two minutes and 
then proceed " under caution," as has to be done when an 
automatic signal is at danger, nor is there any need when 
carrying a tablet or electrical train-staff to send out a flagman 
either forward or in the rear unless "permissive" working is 
in operation, when a flagman in the rear is necessary. 

But the chief trouble of American single tracks is the out- 
lying crossing place. This will be laid out like fig. 162. A and D 
are two signal-boxes or towers which may be 20 miles apart 
with a heavy gradient up which trains struggle between 
them, and on the summit is a passing place B.C., a mile 
or so in length. This may be miles from any habitation, and 
to station an operator there would probably lead to his being 
driven crazy by loneliness or frozen to death by the intense 
cold experienced on some mountains, especially at night. 
The two lines at the passing place must be capable of being 
used in either direction, as the trains that have to pass each 
other may both be going in the same direction, as for in- 
stance a faster having to pass a slower. Here then are some 
of the possible conditions : 

(i). A train from A to D waiting on the west loop at C 

i Mile 


West Loop 



nailing a piece of single line. One scheme was by automatic 
signals and the other by the electrical train-staff. The latter 
would cost one-third of the former to instal, but as operators 
would be required the savings on the initial outlay would be 
swallowed up in three years. The official referred to above 
was not aware of the automatic tablet machine described in 
fig. 179, and which can be worked by the engine-man, guard 
or conductor, and therefore does not require special operators. 

Automatic signals have their advantages. The sections 
can be as short or as long as the traffic demands ; they are 
always on duty day, night and Sundays without any addi- 
tional pay; they are reliable and quick of operation, and can 
be fixed in inaccessible places, such as in deserts, on moun- 
tains, in tunnels and on bridges where operators cannot con- 
veniently be stationed. Finally, they bring with them all the 
advantages that come from " Track-Circuits " and are gen- 
erally interlocked with the siding and. other points in the 

Flectrical methods provide a definite, tangible proof that 
an engine-man is in possession of the section ; but whilst 
they, except in the instrument, fig. 179 above referred to, 
require an operator, which adds to the cost, such operators 
are available for other duties, and possess brains, which is an 
advantage in case of an accident, and can regulate the order 
of the trains, and shunt those that are necessary. An 
operator can hold trains back at the box in the rear if they 
cannot proceed past his box, or which he cannot accommo- 

*The Kail-way Age (Chicago), 28lh April, 1905. 

Easf Loop f* p 

Fig 162. 

for a train to come from D and which has not arrived. 

(2). A train from A to D waiting on the west loop at C 
for a train to come from D- A train from D has been waiting 
at B on the east loop for the arrival of the A to D train, and 
the former must proceed towards and arrive at A before the 
third train can leave D. 

(3). A train from A to D waiting on the east loop at C 
until a faster train from A to D has passed. 

These examples might be multiplied, but sufficient has 
been imagined to allow for the enquiry to be made : How, 
if there be no signalman or operators at B C, can the move- 
ments of these trains be safely governed, not only as to their 
stopping and departure, but the roads on which they are 
to run. 

At the present moment reliance is placed on the orders 
given by the train-despatcher through the operators at A and 
D. They tell the trainmen as to his waiting at B or C and 
what to wait for. But it is easy to imagine a train being told 
to wait at C for four north-bound trains to pass, and for the 
train to start out after the passage of only three, as happened 
on the Western Maryland RR. in the summer of 1905, with 
the result that a collision occurred, and 26 people were killed. 

The Author, when this problem came before him, en- 
deavoured to meet the conditions by the use of one instru- 
ment for through trains from A to D and separate automatic 
instruments (similar to fig. 179) for the line from A to B and 
from C to D f" 1 ' those trains that have to stop in the 



loop. The instruments to be interlocked so that when 

through tablet was out neither of the automatic instruments 

could be used and vice versa. This would, however, mean 

that if a train from D were waiting at B for a train from A 

to D, with a through tablet, to pass, the former could not 

leave B until the latter had arrived at D, nor could a train 

leave A with a through tablet until one coming from D with 

a D-C tablet had arrived at C- Further, two trains that 

had to pass one another would have to stop, one at B and the 

other at C, to deposit the tablet in the instrument. This 

might be met by the brakesman or conductor riding on the 

engine up to the instrument and then alighting and depositing 

the tablet whilst the train was travelling, and then rejoining 

the brake van. The Author feels satisfied that, if the delays 

are not an objection, there is groundwork in his suggestion 

for securing greater safety for operating these outlying loops 

than can be obtained by any other system as long as it is 

impracticable to provide operators. 

The provision of automatic signals is a great additional 
safeguard, but, if the trainmen do not work to their orders con- 
fusion will arise, as the first of two opposing trains to ap- 
proach a section gets possession of it. However, confusion 
is better than disaster, and delay than death. 

Where the electrical tablet or electrical train-staff is in- 
stalled there is no need for the train-despatcher to be com- 
municated with nor for him to ring up the operator. This 
only need be done when the schedule has to be altered. This 
cannot but save time. The system of controlling trains by 
means of orders from the train-despatcher is very flexible, 
but these electrical methods are still more so, as it is easy 
to change the schedule and to send a train forward or to 
keep it back. 

No scheme for controlling single lines by the united action 
of the signalmen at both ends of the section by 
means of block instruments only should be enter- 
tained, as it is so easy for the man at the far end to say that 

the train had arrived when it had not. If, in order to meet 
this, there be provided an electrical contact which has to 
be struck or depressed by the train before the "train-arrived" 
signal can be sent, there is a possibility, should a train 
break loose, for the contact to be made by the first portion 
and for the line to be cleared while it is still occupied. To 
meet this the contact should be made by the last vehicle, but 
the railway companies are still waiting for some method by 
which this can be properly achieved under all conditions. 

The Author has dealt fully with the question of controlling 
single lines in America with the object of informing his 
readers of the methods employed, the difficulties of the situa- 
tion, and the peculiar conditions which have retarded in 
America the provision of those safeguards which are familiar 
in Great Britain, and the absence of which contributes very 
materially to the large number of accidents on American rail- 
roads. There are many in America who see in the block 
system the remedy for most of these accidents. In that con- 
clusion they are right, but those who know the subject must 
admit that before automatic signals, or either of the elec- 
trical methods can be installed on most American single lines, 
there are points to be considered and dealt with that require 
careful meditation, and therefore the delay that is occurring 
in the spread of these safeguards may, the Author considers, 
be justified. 

From a return published in The Railway Age of May i/th, 
1907, it appears that of the single lines in America there were 
or. Dec. 3ist, 1906, 34,493 miles protected by the ordinary- 
block system, 1,088 by " Lock-and-Block " or " Manual- 
Controlled," 2,750 by automatic signals, and 213 by the elec- 
trical train-staff. From the annual report of the Interstate 
Commerce Commiss'on for 1906 the Author gathers that out 
of 222,340 route miles of railway 201,358 are single, so that 
only 38,564, or 19 per cent., are protected by any means of 
" securing an adequate interval of space between following 




Tycr's No. 5 Tablet Machine. 

This machine is very simple in construction and con- 
venient in size. The only objection that can be urged against 
it is that it does not give visual indication to a signalman 
as to whether he has given, or obtained, permission for the 
withdrawal of a tablet, but there is nothing serious in this, 
as, providing the signalman acts up to his duty, his Train 
Register book will give him all the information he may re- 
quire. Certain it is that if he wants a tablet and ought not 
to have one, he will not be able to get it. 

Fig. 163 illustrates the No. 5 machine, the lower part 
being the instrument and the upper part the bell, which is, 
however, separate. The whole of the mechanism is enclosed 
within the casing T. 

Through the cover of the casing passes a vertical axis 
T 1 with a milled head, by which it can be turned by hand in 
either direction. There is also in the cover a slot of sufficient 
size to admit a tablet inserted edgeways, with notches cut in 
the slot to admit the fingers when taking out or inserting a 
tablet. This slot is covered by a hinged lid T 3 , which can 
be turned back to give access to the slot. On the axis T 1 
within the casing is fixed a disc D (fig. 164), having a number 
of radial slots D 1 , each capable of admitting a tablet (T 2 ) 
inserted edgeways, and below this disc, also on the axis T 1 , 
is a wheel W which has projecting from its face a circular 
bead W 1 fitting the notch shown in the edge of the tablet in 
.fig. 165, so that when the tablet is inserted its lower edge 
rests on W, with its notch enclosing the bead W 1 . On the 
upper face of the wheel W project ratchet teeth W 2 , en- 
gaged by a dropping pawl W 3 . Also at the edge of the 
wheel W there are ratchet teeth W 4 engaged by a spring 
pawl XV 5 . Both sets of ratchet teeth correspond in number 
with the slots D 1 of the disc D, and the teeth are sloped in 
opposite directions, so that one pawl, W 5 , prevents the axis 
T 1 , the disc D, and the wheel W from being turned for- 
wards, that is in the direction of the hands of a clock; the 
other pawl, W 3 , prevents them from turning backwards. 
The pawl W 5 is cranked upwards and has projecting hori- 
zontally from it a horn W 6 which, coming in front of that 
tablet T 2 which may happen to be in the neighbouring slot 
of the disc D, prevents the disc and wheel from turning 

backwards, even if the pawl XX' 3 should leave the wheel free 
to turn. The disc D also has on the upper face of the boss, 
near the centre, teeth D 2 corresponding in number with the 
slots D 1 , and the lid T 3 is made with a tail which, when the 
lid is opened, engaging between a pair of the teeth D 2 , pre- 
vent the disc D, wheel \V, and axis T 1 from being turned in 
either direction until the lid T 3 is closed. The spring pawl 
W 5 has a piece of insulating material, W 7 , bearing against 
a contact spring W 8 , so that when the pawl is pushed back 
so as to be disengaged from the teeth XV 4 the spring W 8 is 
moved out of contact. The dropping pawl XX" 3 (fig. 1643 for 
enlarged view) is one arm of a bell crank lever, the other 
arm XX' 9 of which comes behind a lug M 2 projecting from 
the armature M 1 of an electro-magnet M, so that when this 
electro-magnet attracts its armature, the pawl XX" 3 is lifted 
out of its engagement with the teeth W 2 . 

In addition to the disc D and the wheel XX r before described, 
the axis 7' 1 carries a multiple cam C which operates as a 

L ill'jill 

Fig. 163. Tyer's No. 5 Tablet Machine 




Fig. 164.7. 



v/i C3 







: -; 

Figs. 164 and 1640. Tver's No. 5 Tablet Machine. 

commutator, having half as many prominences C 1 and half 
as many hollows C 2 between them as there are slots D 1 or 
teeth \Y 2 or \Y 4 . A roller C 3 on a lever C 4 urged by a spring 
bears against the edge of the wheel C so that as the wheel 
revolves the lever C 4 is moved to and fro, moving contact 
springs C 5 , C 6 and C 7 linked to the lever. 

The movements for issuing a tablet are as follows : 
The signalman at A, having to send a train to B, would 
depress his plunger S 1 , which would sound the gong G in 

B, and B, having acknowledged it and received the pre- 
scribed signal, can, by depressing his plunger S 2 whilst A 
depresses his plunger S 3 , cause a current to pass from the 
local battery P at A (fig. 163) through the coil of the magnet 
M. The armature M 1 being attracted raises the pawl W 3 , 
enabling the wheel W to be turned backwards, but only one 
tooth, because the roller C 3 is moved by the partial revolu- 
tion of C so as to break a contact at C 5 , thus opening the 
circuit of the magnet M, whereupon the pawl \V 3 falls, 

o 2 






'3 C 1 


r > 


Fig. 165. Tyer's No. 5 Tablet Machine. 

stopping the further movement of the wheel W. But the 
wheel having made one movement has brought a tablet to 
a position under the lid T 3 , and the tablet in its passage has 
pushed away the upper part of the pawl \V 5 and broken the 
contact of W 8 , thus cutting the local battery P out of circuit 
with C 6 . On opening the lid T 3 the tail of the lid engages 
with the teeth D 2 and prevents the axis T 1 being turned, and 
on the tablet being withdrawn the pawl W 5 again engages 
with W 4 , and contact with W 8 is re-established. 

Should B have withdrawn a tablet, on its receipt at A, 
the latter can open the lid T 3 and insert the tablet, but by so 
doing he pushes away \V 5 , breaking the contact at W 8 , 

which is not re-established until A, having closed the lid, 
turns his wheel one tooth forward, thereby removing the 
tablet to position shown by 7' 4 (fig. 1640), and thus allowing 
W 5 to engage with the next of the teeth W 4 and W 8 to again 
make contact. 

On the upper part' of the axis T 1 is provided an index T 5 
pointing to graduations on the flange T 6 so as to show the 
operator the number of tablets in the instruments. This is 
necessary, as it often happens that the traffic in one direc- 
tion is greater than in the other, so that the tablets gradually 
find their way to one end, and when, this occurs the tele- 
graph lineman has to be sent for to regulate the supply. 

\^s r 







Fig. 166. Tyer's No. 6 Tablet Machine. 




tlHC | 



Fig. 167. Electrical Connections for No. 6 Tablet Machine. 

Tyer's No. 6 Machine. 

This is illustrated by figs. 166 to 171. In this instrument 
are provided discs for indicating similar to the block 
what signals the signalman has given and received. The 
discs are made to show " In " and " Old." \Yhen the 
upper disc shows "In" it indicates that the down line is 
clear, and when " Out " it indicates that permission has 
been given to the signalman at the other end of the section 
to withdraw a tablet for a down train, and that the other 
man has accepted the permission and withdrawn the tablet. 
The lower disc indicates to the signalman when at " Out 
that he himself has sought and obtained permission to with- 
draw a tablet for an up train. 

The tablets are contained in the case F, and to gain 
access to them, or to return a tablet to the instrument, the 
slide B has to be withdrawn, and it is this slide which is 
electrically controlled from the other box. The galvanometer 
needle C indicates the passage of the electrical signal 
between the two signal-boxes when the man at the other 
box is holding down his plunger (as per code) for the pur- 

pose of unlocking the slide. It also indicates all preparatory 
signals, describing the type of train, departure of same, &r. 
The plunger D is for ringing the bell signals as per code, 
and the commutator A for unlocking and locking the slide. 
The switch E is a mechanical arrangement for placing nn 
inserted tablet into the cylinder in the case F. 

In the apparatus a slide, operated by hand, is employed 
for moving a tablet out of, or into, the receptacle provided, 
but the movement of this slide is governed by pawls which 
can only be released by the operation of electro-magnets 
having their coils suitably connected to a relay operated by 
electrical currents transmitted by the line-wire connecting 
the stations. One object of this invention is to effect the 
movements of these pawls by the hand of the operator, em- 
ploying the electro-magnets merely to withdraw catches so 
as to leave the pawls free to be moved. Thus, pawls of 
greater weight and therefore of more certain action can be 
employed than when they themselves have to be moved elec- 
trically. The catches above referred to can only be released 
by the joint operation of the signalmen at the two stations, 
and thus all danger of a tablet being improperly issued is 

The apparatus is provided with a commutator and sig- 
nalling plunger so arranged that the signalmen at the two 
stations, A and B, can operate in the following manner :~- 

Assuming that A desires to issue a tablet for a train pro- 
ceeding towards B, A by depressing his signalling plunger 
gives B the preconcerted signal. B, by his plunger, can 
answer, and then, if his electrical connections as well as 
those at A be in suitable condition, he can transmit a current 
whereby a catch in the apparatus at A is withdrawn, enabling 
A, by turning his commutator, to release a pawl and then 
to draw out his slide and issue a tablet. By this act of 
releasing one pawl A causes the engagement of another 
pawl which prevents his slide from being pushed in again, 
until B, having received the tablet, shall have put it into his 
apparatus, thereby so acting on his own commutator that he 
can send to A a current of the opposite sign to that which 
released the slide for drawing it out, thus causing with- 
drawal of a catch enabling A to release the second pawl and 
push in his slide. All the connections are thus brought into 
condition for operating again in like manner. 

Sometimes tablets are issued for shunting purposes, and 
thus A may receive back the tablet which he has issued with- 
out its passing to B- Although A having drawn out his slide 
to issue this tablet cannot push his slide in again while it is 
empty he can, by re-inserting the tablet, cause the pawl to 
be released, so enabling him to push in the slide with the 
tablet in it, and to restore his commutator to condition for 
operating afresh. 

The lever at the side of the instrument for raising tablets 
out of the recess of the slide becomes bolted when the 
operator moves his commutator for the purpose of taking 
out a tablet. The sliding bar which in Tyer's earlier appa- 
ratus was pushed by the edge of a tablet when the slide is 
put back and thereby acted on the electrical connection, has 
its front end, in the present instrument, jointed and counter- 



Fig. 168 Tyer's No. 6 Tablet Machine. 

poised by a weight or spring so as to yield to the weight of 
superimposed tablets above it, but to present its end directly 
to the edge of an entering tablet so as to be pushed by it. 

Fig. 166 is a view of one complete set of the apparatus, 
consisting of the tablet receiving and issuing mechanism 
B, the bell signalling apparatus D, and the electrical com- 
municating apparatus A; fig. 168 is a side elevation and fig. 
171 is a plan of the tablet receiving and issuing gear within 
the casing A; figs. 169 and 170 are part longitudinal sec- 
tions of the same; fig. 167 is a diagram showing the elec- 
trical connections. 

The cylinder D, containing the tablets D 1 , and the slide 
D 2 , by which a tablet can be taken out or put in, are ar- 
ranged as seen, as also the hand lever D 5 for raising the 
tablets in the cylinder, but this lever is, in the present in- 
strument, prevented from rising to its full height, sufficient 
to raise the tablets out of the slide, by the nose d coming 
against a piece d 1 fixed on a rod d 2 , which is moved by a bell 
crank d 3 , acted on by a rod d 4 , which is linked to one of two 
arms on the axis T of the commutator. The slide D 2 is, as 
before, stopped by a pawl E 2 , which prevents it from being 
drawn out. There is also a pawl e, which prevents the slide 
P 2 from being pushed quite in. This latter pawl is re- 

leased by the descent of a rod e 1 , which is linked to the other 
of the two arms on the axis T of the commutator. 
The rods d* and e 1 have on their lower parts shoulders, over 
which project the armatures d 5 and e 2 respectively. The axis 
of the commutator cannot be turned in either direction unless 
one or other of these armatures is attracted. When e 1 is 
attracted the commutator can be turned back to the right. 

By turning the commutator to the left the rod d* is 
lowered, depressing one arm of the bell crank d 3 , and thereby 
causing a projection d 6 to act on the tail of the pawl E 2 , 
raising it out of the notch of the slide D 2 and so permitting 
the slide to be drawn out with a tablet. 

By turning the commutator to the right the pawl E 2 
becomes again free to engage in the notch of the slide, and 
the rod e 1 is lowered, depressing one end of the pawl lever e, 
so that the pawl cannot engage in the notch of the slide D 2 . 
On the axis T of the commutator, which can be turned by 
the knob T 1 when the armature d 5 and e 1 permit, is fitted a 
disc T 2 free to revolve a certain distance as determined by a 
slot in it engaged by a pin on the arm to which d 6 is linked. 
Also rigidly connected to this disc is another disc T 3 of insu- 
lating material with two metal semi-circles on its face, insu- 
lated from one another. These metal semi-circles, when the 



Fig. 171. Tver's No. 6 Tablet Machine. 

signalling plunger T* is pushed in, make electrical contact When the slide D 2 is pushed in with a tablet the sliding 

with the four springs t 1 t- t 3 t 4 connecting them in pairs in bar G is pushed back, causing the roller t 6 and rod i 5 to rise, 
circuit, according as the commutator axis T is turned to the By the rising of the rod t 5 acting on a pin t 7 on the disc T 5 
right or left. A rod t 5 carries at its lower end a roller i 6 , this disc is turned partly round, turning with it the commu- 
\\hich rests on an incline g on the sliding bar G. tating disc T 3 and thereby altering the circuits, should the 



plunger T 4 be pushed in. But, when the armature e 4 is 
attracted, an arm e 5 extending up from it bears against a pawl 
e s which held the disc T, releasing this pawl and allowing 
the disc T 2 to be moved back by the weight e 7 . Z is a bell 
crank, the upright arm of which is hooked so as to engage 
on a pin z and thereby prevent the rod c l from being raised 
should the slide D 2 not have been pushed in. But when it 
is pushed in, a projection z 1 on the slide raises the lower 
arm of the crank Z, thus moving its hooked arm clear of the 
pin 0. 

It will be noticed that the attraction of the armatures d 5 
and e 2 has not to perform the work of releasing the pawls 
which hold the slide D 2 . All that these armatures have to do 
is to permit the signalman to turn the commutator axis T, 
which, in turning, effects two purposes : Firstly, it releases 
the pawl which is holding the slide D 2 , and, secondly, it puts 
the disc T 3 in such a position that, when the plunger T 4 is 
pushed in, it makes the right contact of the springs t l 2 3 i 4 . 

Tyer's Automatic Tablet Machine. 

In addition to the two machines just described, Mr. Tyer 
has introduced another which is known as the " automatic " 
and which is similar to the No. 5 and No. 6. This machine 
also has no visual indications. It is illustrated by figs. 17 
to 1733 and Mr. Tyer claims for it that whilst it is quite as 
safe and equally as effective as any other machine, it has 
fewer parts and can be supplied at a lower cost than any 
other tablet machine. 

Fig-. 172. Tyer's Automatic Tablet Machine without Visual Indications. 

Fig. 172 gives a view of the machine, the notable feature 
of which is the hopper B on the top. Instead of a tablet 
being withdrawn and restored by means of the slide E, the 
tablet is only taken out by that means, but it is restored to 
the instrument through the hopper B. 

The mechanism is illustrated by figs. 173 and 1733. A 
tablet can be introduced into the slot of the hopper B as far 
as permitted by a stop h 2 and then it can be turned partly 
round by bringing its slot in line with the inclined way b 3 
down which the tablet slides to the receptacle C in which a 
number of the tablets can be accommodated. The tablet in 
sliding down the incline b 3 moves aside a counter-weighted 
lever D, carrying a pawl d, which, on the return stroke of 
D, effected by the counter-weight, turns a ratchet wheel d 1 
one tooth round, thereby turning partly round a commuta- 
ting wheel d 2 , against which bears the finger d 3 of a weighted 

bell crank lever. This lever carries insulated springs d 4 , 
which, when the parts are in the position shown in fig. 173, 
bear against upper contacts d 5 , and when they are as shown 
by fig. i73a bear against lower contacts d 6 . 

Under the receptacle C is arranged the horizontally 
sliding plate E, having a recess c of such depth as to receive 
one tablet T and no more. When pushed in as in fig. 173 
it is held by a hook pawl e 1 attached to the armature of an 
electro-magnet e 2 , so that the slide E cannot be drawn out. 
But when a current excites the magnet e 2 , as determined by 
the action of the signalman at the other station, its armature 
is attracted, raising the hook e 1 and leaving the slide E free 
to be drawn out along with a tablet in its recess e. On the 
slide E is fixed a bracket F having two laterally projecting 

Tver's Automatic Tablet Machine without Visual Indications. 

pins / f 1 , between which is situated an arm of a lever G, 
carrying a pawl g, which acts on a ratchet wheel g l on the 
same axis with the ratchet wheel d 1 and with the commu- 
tating wheel d 2 . When the slide E is pulled out to with- 
draw a tablet, the pin / moves the lever G, so that its pawl 
turns the commutating wheel partly round, altering the con- 
tacts at d 5 and d 6 . When the slide E is pushed in, to be again 
held by the hook e 1 , the pin f 1 pushes back the lever G, 
causing its pawl to take a position ready to act on another 
tooth of the ratchet wheel d 1 . Thus when a tablet is put in at 
B and slides down to C the commutating wheel d 2 , as already 
explained, is turned so as to alter the contacts d 5 and d 6 , and 
also when the slide E is drawn out to take out a tablet, the com- 
mutating wheel is turned so as to alter these contacts. Thus 
the introduction of a tablet, or the removal of a tablet, alters 
through the contacts d 5 and d 6 the current communicating 
with the next station from + to and then from to + 
and so on alternately. Under the slide E is pivotted a coun- 


ter-weighted lever H, the front end h of which, if the slide 
were partly pushed in with a tablet in its recess e, would 
catch the edge of the tablet and so prevent the slide from 
being pushed further in. Thus the slide cannot be pushed 
home with a tablet in its recess. When the slide E without a 
tablet is pushed home, it makes contact by a bridge piece 
with two spring pins at K, thus establishing electrical com- 
munication by the line ; but when the slide is drawn out or 
pushed only partially in there is no contact at K, and conse- 
quently there can be no communication by the line wire. A 
tappet L, projecting down from the slide E, meets at each 
end of the stroke, stops / / which determine the stroke of the 
slide, and are preferably faced with wood or other deadening 
material. The gauge G in fig. 172 is provided to indicate 
how many tablets are remaining in the cylinder. 

Tyer's Automatic Machine with Visual Indications. 

Another cheaper form of instrument is that illustrated by 
fig. 174, the working of which is as follows : 

Take the two signal boxes at the ends of a single line sec- 
tion as A and B and imagine that A has a train ready to go to 
B. After giving the usual signals, as per code, A holds down 
his bell plunger. 

Gong arid Relay 

Fig- 174- 

Tver's Automatic Machine, 
with Visual Indications. 



B depresses his switch-plunger with his left hand, and at 
same time withdraws his slide S with his right hand; the 
withdrawal of this slide (which can only be half way) will 
reverse the commutator in the instrument, and will bring the 
signal " Up-train-approaching: " in view. B will then depress 
his bell-plunger, holding down on same for a few seconds. 
A, upon receipt of this signal, will hold down his switch- 
plunger with his left hand, and with his right hand draw out 
his slide marked S to its full extent, which will bring in view 
the signal " I' p-t rain-on-line." 

A will remove the tablet from the recess in the slide, and 
will hand the same to the engine driver. A then gives the 
departure signal to B- 

As soon as the train arrives at B, the man there 

withdraws the slide R, inserts the tablet, pushes the 
slide home, the passage of the tablet into the cylinder 
unlocks his instrument, and B then pushes in the slide S, 
which restores his visual signal to " Line-closed." He will 
then give A the "Arrival " signal, holding down on the last 

A, upon receipt of the "Arrival " signal, will depress his 
switch plunger, and push home his slide, which will, at the 
same time, restore his visual signal to " Line-closed." 

For a tablet, that has been taken out for shunting, or for 
the purpose of entering some intermediate siding, and which 
is returned to the station from whence it was issued, the 
manipulation of the machine is practically the same ; the man 
who issued the tablet merely returns the same into his 
machine by means of the slide R, which will restore the 
apparatus after the passing of the usual signals. 

The signalmen at each end of the section have at all 
times an absolute record of what has been done on both 
instruments, which record cannot be disturbed until the tab- 
let has been placed in the apparatus at either one end or the 

The screens with the visual signal indications are lettered 
in pairs, and the screen at A would be lettered " Up-train-on- 
linc " on red ground as shown, but the screen at B would be 
" Down-train-on-line." The normal condition of both instru- 
ments at A and B is " Line-closed." 

Tyer's No. 5 Permissive Tablet Instrument. 

All the machines so far described are for "Absolute " 
working only, i.e., only one tablet for a section can be out at 
the same time. But the Xo. 5 machine (fig. 163) has been 
improved and adapted for " Permissive " working, by which 
is meant that a second and subsequent tablets can be with- 
drawn from the machine, although the first has not been 
placed in the instrument at the other end of the section. 
Such working is not in use in Great Britain, but it has, of 
necessity, to be adopted in North and South America, India, 
Africa, Australia, and other countries where there are 
long distances between stations, and consequently it 
would cause too great a delay to a second train for it to 
wait until the first train had passed out of the section, so 
DOWN TRAIN ON LINE V provision must be made for issuing subsequent tablets 
out of the same machine for following trains. This prac- 
tice is safeguarded by preventing a tablet being with- 




drawn at the other end for a train to come in the opposite 
direction until all the trains for which tablets have been issued 
have delivered up their tablets, and they have been inserted 
in the machine at the other end. 

On referring to figs. 163-4 " l w '" be seen that the tablets 
revolve on a wheel, access to which is obtained by means ol 
a lid in the upper part of the case, the tablets being replaced 
in the machine through the same aperture. 

The lid is marked T 3 in fig. 175, and a part section of a 
tablet is shown below. Part of the modification lies in the 
addition of the slide e in the lower part of the instrument, 
a section of which is given in fig. 178. When more than 
one train has to pass through the section (say A to B), the 



Tyer's No. 5 " Permissive " Tablet Instrument, 

Fig. 17$ 

necessary bell signals are passed between the two boxes and 
B receives permission to withdraw a tablet. In the lower 
part of the instrument there is a bolt h, one object of which 
is to prevent the slide e being fully pulled out, so that a 
tablet can be inserted in the recess g, unless the operation 
of pressing down the rod b be carried out. The rod b 
presses on one end of the bolt h and raises it, so that the 
slide e can be pulled fully out for the signalman to insert the 
tablet, as illustrated by fig. 176. The slide is pushed in again 
and the tablet comes against the slide i and forces the slide 
back. The signalman then presses the rod b further down, 
so that it passes through a hole in the slide i into a hole in 
the base plate of the apparatus, so that the slide e is held and 
cannot be withdrawn. In pressing down the bolt b the key 
on its boss d is withdrawn from the boss of D, the wheel 
carrying the remainder of the tablets, so that the cam wheel 
and commutator C can be turned without moving the wheel 
D. A then sends to B a current which frees the commutator 
C and allows it to electrically coincide with the instrument 
at A and allows A to take out a tablet for the first train. In 
doing this the key on the boss d is moved away from the slot 

in D, consequently neither d nor b can be raised, and 
so the tablet in the slide e is held secure. A in due 
course will ask permission to send a second train, 
which will allow B to turn his commutating cam wheel 
C one further division, so as to once more electrically 
coincide with A, and then B can give permission for a 
second tablet to be withdrawn. This can go on as 
long as necessary, each permission taking commu- 
tating wheel C a division further. As the trains, how- 
ever, arrive at B and the tablets are inserted into the 
instrument there, the wheel C gradually revolves back 
to its normal position, until, the last tablet being in- 
serted, the key of d. and the slot of D become lineable, 
and the rod b is then forced upwards by the spring c, 
and the key of d again engages in the slot of D, so 
that C and D are again coupled, and the slide e is 
released, so that B can draw it out, and, taking out 
the original tablet, he restores it to the instrument, and 
then the communication is once more complete, and 
either man is free to ask permission to send trains 
from either end. 

Tyer's Absolute Automatic Tablet 


It was stated in the preceding chapter that one of 
the objections raised by American railway officials 
against the British methods of working single lines by 
the electrical tablet or electrical train-staff is that its 
adoption necessitates the provision of operators to 
manipulate the instruments and to work the signals. 

Messrs. Tyer & Co., Ltd., have an instrument which 
should interest not only American officials, but those in 
India, Africa, South America, &c. It gives all the safe- 
(_) guards obtained by the ordinary tablet machine, but it 
does not need special operators, as the instrument can 
be worked by the driver, guard, or any other trainman. 
By this means the establishment at stations can be 
reduced to a minimum, and at passing-places where 
the points are self-acting no staff is required at all. This is 
an advantage, too, in those unhealthy districts where malarial 
and other fevers are prevalent and where constant changes of 
staff are necessary. 

The machine has all the advantages of the other tablet 
machines in that only one tablet for a section can be out at 
the same time; that a tablet can be withdrawn from either 
end; that a tablet can be returned to the saipe instrument if 
it should be found that it is not required, or for any other 
reason, and that a train need not pass through the section 
to get rid of the tablet, but can return to that end of the sec- 
tion it entered at, and restore the tablet to the same machine. 

The instrument is illustrated by fig. 179. On arrival at a 
tablet station the guard or driver will go to the instrument, 
where by means of the indicator he will see whether a tablet 
is out for the section ahead. If the indicator is central, as 
shown, then no tablet is out, but if it points to either of the 
two indications given he will know whether it be for a train 
in front of him or one that is coming towards him. If his be 
an " up " train and the indication points to " down-f rain-on- 



Fig. 179. Tver's Absolute Automatic Tablet Instrument. 

line " the man knows that it is useless to try the machine, 
and he simply waits until the other train arrives. If the indi- 
cator points to " up-t rain-on-line " then he waits till the 
needle falls to normal. 

The instruments normally contain 12 tablets, but there is 
room for 24. These rest one on each other in the front of 
the'case as seen through the screen. To obtain a tablet the 
lower drawer is drawn forward, which brings out a tablet 
lying in an aperture in the drawer. There is, of course, only 
room in the aperture for one tablet; but before the drawer 
can be pulled out the plunger has to be pressed in. This 
allows an incoming current from the instrument at the other 
end of the section (say B) to pass into the instrument at A, 
which closes the line relay and operates a lock on the drawer. 
The drawer can then be withdrawn and the indicator at A and 
at B turn to " up-t rain-on-line," and the plungers of both 
instruments then become locked so that a second tablet can- 
not be taken out of either instrument. On arrival at the 
other end of the section (at B) the guard gets possession of 
the tablet from the driver and proceeds to the corresponding 
instrument, and, pulling out the upper drawer, places the 
tablet in a recess and restores the drawer. On the return of 
the drawer the tablet falls into the machine, and in its descent 
the line is automatically broken down so that both instru- 
ments are restored to normal and a train can be sent from 
either end. At the same time the needles of both instru- 
ments fall to the vertical position. 

McKenzie and Holland's Tablet Instrument. 


Messrs. McKenzie & Holland, Ltd., supply a 
instrument similar to that illustrated by fig. 180. 

In the upper part there is the usual screen a, wherein are 
given the following signals : '' Tablet-otit-up-line," " Tab- 
let-out-dovn-line," " Clear " and a plain green disc. Below 

this is the galvanometer and then the ringing key, b (since 
replaced by a plunging key). 

The long slot, c, in the lower part is to show to the sig- 
nalman how many tablets are in the instrument. Below the 
slot is the lower drawer, the handle-knob, d, of which passes 
along the slot of a lid over the drawer. The smaller knob, e, 
belongs to the lid. 

Fig. 180. 
McKenzie and Holland's Tablet Instrument. 

\Mien permission has been received to withdraw a tablet 
the lower drawer is pulled out. \Yhen a tablet has to be in- 
serted in the instrument the upper drawer, /, above the slot 
is pulled out. In the illustration it is shown out, and the 
handle, g, which is hinged, hangs down out of the way. The 
upper smaller knob, h, belongs to the lid over the upper or 
receiving drawer. 

Switching-out Tablet Stations. 

One of the difficulties in the working of single lines is 
that at night, so long as any trains are liable to come, the 
signal boxes at crossing places have to be kept open so that 
tablets may be exchanged and the points set and signals 
lowered for the train to pass through the crossing places on 
the right road. This, however, means expense, as men may 
remain on duty for hours and have nothing to do. 

The Caledonian R. Co. provided for this difficulty for 
their early morning mail train to Oban, on the Callender and 
Oban line, by each man, after the last evening train had 



passed, setting the down road and withdrawing a down 
tablet, which was left in the signal box. When the Oban 
train arrived in the early hours, the guard went into the 
signal box (say B), and left the tablet the train had brought 
from A, picked up the tablet lying there which allowed him 
to proceed to C, at C he left the B-C tablet and took one for 
C-D which had been left by the signalman at C. Then when 
the respective signalmen came on duty in the morning, they 
had to place the tablets they found lying in their boxes into 
the instruments in order to make the communication good. 

This method of working is satisfactory when one train is 
expected, and it is known in which direction it will travel, 
but when trains may come in either direction and at uncer- 
tain hours, this cannot, of course, be done. 

The method employed by some companies in order to 
switch out a signal box at night or on Sunday is to have a 
separate pair of instruments to cover the night section. 

If there be, during the day, two sections A-B and B-C 
and B has to be closed at night, there would be a separate 
pair of instruments for A-C, the tablets of which would not 
fit the A-B nor B-C instruments, in fact, they might be square 
instead of round, as is usual. When B had to be closed the 
ordinary instruments would, after the necessary exchange of 
telegraphic signals, be switched out and the A-C switched in. 

If B were a crossing place the signalling would be ar- 
ranged as shown by fig. 181. All the traffic would have to 
pass over the up line. No. 4 points would be normal, No. 6 
" over," and this would free signals Nos. i, 2, 7, 8, which, 

Fig. 181. Closing Tablet Station. 


when lowered, would free lever No. 5. On this lever being 
pulled over it would back-lock those levers and allow for the 
A-C instruments to be switched in. When this was done 
No. 5 lever would be electrically locked in its over position. 

Messrs. McKenzie & Holland, Ltd., have an arrange- 
ment whereby tablet stations can be switched out. It is in 
use on the South Western, Cambrian and Highland Rs. 

The two sections being A-B and B-C, and B the box to 
be closed, there are the usual instruments for A-B and B-C 
and special instruments A-C when B is closed. There are also 
three switch instruments, one in each box, those in A and C 
being similar. 

B ns 

Fig. 182. 

13, Interlocking Lever. 14 used for long section working. To set road for long section 
working -.full over ij half-way, 6, Q, lo, 3, 2, 14, 77, 13 second half. 

The signalling at B is arranged as in fig. 182, but the 
special lever, No. 13, is provided with two locks like those in 
fig. 183. The two keys that fit these locks correspond with 
the locks on the switch instrument at B. One of the keys is 
long and the other short, and when normal working is in 

operation, i.e., with short sections, then the "short" key 
is out of the upper lock on the lever and in the " short sec- 
tion " lock of the switch, so that the ordinary tablet instru- 
ments are in circuit, and the special lever is held normal by 
the " short " key being out of the lock. When the signal 
box is to be closed A and C give B permission, and he with- 
draws the " short " key from the switch, inserts it in the 
lock and pulls the lever half-way. This back-locks the 
" short " key in its lock and frees the interlocking that 
allows the road and signals to be set to admit for through 
working on one line. When the necessary point and signal 
levers are in their proper position the special lever can be 



Fig. 183. 

pulled fully over, and this frees the " long " key in the lower 
lock. This key is put in the " long " lock of the switch, and 
being turned allows for the "long" section instruments to 
be joined up. The men on either side turn their switches to 
correspond, and this locks up the " long " key, and therefore 
the special lever cannot be moved and the road and signals 
are held to permit through working. 

Banking Engines on Single Lines. 

Before leaving the question of tablet working a descrip- 
tion of an arrangement in use at Oban may be of interest. 

On leaving Oban there is the heavy Glencruican bank to 
climb, and it is necessary for certain trains to be assisted in 
the rear. The length of the tablet section extends from Oban 
to Connel Ferry, a distance of over 6 miles, and as the bank 
engine has no need to go further than the summit 3 miles 
out a difficulty arose as to the security of the section until 
the bank engine returned. It was a waste of power and time 
to let the bank engine go through the section and then 
return, whilst on the other hand it was not safe to allow the 
bank engine to be in the section, so eventually the arrange- 
ment illustrated by fig. 184 was designed. 

This is one of the ordinary wooden train staffs enclosed 
in a box, through which pass the wires to the tablet instru- 
ment. The train carries, as usual, the tablet, and the bank 
c engine carries the staff, and the apparatus is 
arranged so that when the staff is withdrawn 
the tablet communication is broken down, and a 
second tablet cannot be obtained until the staff has 
been restored. The tablet can be inserted at the other end, 
and as the bell communication is not broken the stations can 
speak to one another, if necessary. 

The illustrations show the staff in position for withdrawal. 
To withdraw the staff it has to be turned from right to left 
into the position shown, and this operation actuates the slide 
a, upon which is fastened a piece of wood c, carrying a brass 



Section A . B 


' Section E. F 


Section G O 

Fig. 184. Dunn and Steven's Staff Box for Banking Engine. 

inclined plane 6. A contact spring d, fitted with a roller e, 
forms contact between the wires leading to the pair of tablet 
instruments (one at Oban and the other at Connel Ferry), 
and when the staff is turned, prior to withdrawal, the slide 
is moved from right to left, and the spring d is so raised off 
the stud /, and contact is broken, severing connection 
between the tablet stations. Also contained in the staff box 
is the lever shown in dotted lines. It is fitted with a pin h, 
for which two recesses are cut in the slide a, in one of which 
the pin h is held by the spring g, and prevents the slide a 
from being moved unless the pin h has been raised out of the 
recess, but this can only be done by the train staff being 
inserted and turned and so raising the lever. 

Tablet Pouch. 

For the protection and preservation of the tablets and to 
facilitate their handling they are generally handed to the 
engine men and carried in a pouch similar to that shown in 
fig. 185. This pouch, however, was patented by Mr. J. A. 
Hoffe, of Cape Town, and is manufactured by Tver & Co., 
Dalston, N. 

Fig. 185. Tablet Pouch. 

This is a pouch b open in front so that the lettering on 
the tablet can be seen and with a leather back d. It is pro- 
vided with a large leather covered ring or hoop a through 
which the men pass their arms when exchanging a tablet at 
speed. A strap / secured to the back passes in front of the 
tablet and through a slot in the stud g. This stud passes 
through the centre hole of the tablet and assists in holding 
it in position. 

Exchanging Tablets and Electrical Train Staffs at Speed. 

In the Author's Mechanical Railway Signalling the prin- 
cipal appliances which are in use for exchanging electric tab- 
lets and train staffs are fully described. 

Amongst these is Manson's, and fig. 186 is a view of the 
apparatus that is fitted to the engine. When running and 
not required it lies against the engine as seen, but is lowered 
to a horizontal position by means of a lever when a tablet has 
to be picked up, dropped or exchanged. 

The tablet to be delivered is placed in a pouch with a pro- 
jecting top which rests on the top of, and the pouch hangs 
between, the two springs at the back. At the signal box is a 
standard, fig. 187, with a similar deliverer and receiver. 
These coincide with those on the engine, so that the engine 
picks up a fresh tablet by means of the jaws seen on the left 
in fig. 186, and similar jaws on the standard receive the 
tablet that has to be given up. Normally the arm of the 
standard is back from the line and has to be pushed out by 
means of the lever seen, fig. 187, in the signalman's hands. 

The apparatus used by the Great \Yestern R. for ex- 
changing electrical train-staffs is described in Mechanical 



Fig. i 86. 

Hanson's Tablet Exchanging Apparatus. 

Fig. 187. 

Railway Signalling and the annexed fig's. 188 and i88a show 
views of the apparatus. 

Whitaker's Exchanger for Train-Staffs and Tablets. 

This is illustrated by figs. 189-191, adapted for exchang- 
ing train-staffs, but it is readily arranged for exchanging 

On the engine, fig. 189, is a combined deliverer and 
receiver. When out of use this is close to the side of the 
engine and is pushed out as shown when required. The 
train-staff to be given up is placed in a rubber pouch with a 
steel ring and carried at the rear end of the apparatus, and 
kept in position by a spring clip. The receiver consists of a 
gunmetal jaw with two triggers in the front and with a 
rubber pad at the back. 

On the line, at the train-staff station, is a standard similar 
to that Illustrated by fig. 190. The jaws of the apparatus on 
the engine engage with the loop of the train-staff .pouch 
carried on the lower arm of the standard, and a similar 
receiver on the upper arm of the standard seizes the loop of 
the train-staff pouch from the engine. The apparatus at the 
moment of exchanging is shown by fg. 191. 

The arms of the standard are normally parallel with the 
running line, and are turned to a right angle with the line by 
the signalman when putting a staff ready for exchange. The 
standard is provided with two bevel wheels as shown, but 
these have teeth on only a quarter of their faces and are pro- 
vided with a stop to prevent them moving further than the 
correct distance. When that point is reached the weighted 

Fig. 188. 

Train-Staff Exchanging Apparatus, Great Western Railway. 

Fig. 188,7. 



lever, see fig. 191, is slightly past the perpendicu'ar. The 
shock given by the receipt of the train-staff into the receiver 
is such' as_ to throw the weight over the centre, and it then 
falls and the bevel wheels turn the standard so that the arms 
are cleared of the running lines. This is a very good feature. 


Fig. 189. 

The apparatus is in use on the Somerset and Dorset Joint 
Line, of which railway the patentee, Mr. \Yhitaker, is loco- 
motive engineer, and it is manufactured by the Railwav 
Signal Co., Ltd. 

\Yhere tablets made of compressed fibre are used no ex- 
change apparatus is necessary, as those tablets weigh less 
than 2 ozs., so that if, in exchanging them, a man gets struck 
on the arm, no injury is done to him. 

That some exchanging apparatus is required is shown by 
the fact that during the year 1905 there were 19 railway 
servants injured in exchanging tablets or train staffs and 20 in 
1906. The general report on railway accidents for the year 
1905 says of these that 

a large proportion were caused by unduly high speed of trains. 
Mechanical appliances by which the risk of exchanging by hand is 
. avoided are already in use, and their extended employment, coupled with 
a limit of speed at places where exchange by hand is continued, will be 
pressed upon the railway companies. 

Unlocking Starting Signal by Tablet or Electrical Train-Staff. 

It is most desirable that the signal for entering a section 
should be unlocked by the tablet or train staff so that it 
cannot be lowered unless and until the tablet or staff is out 
for the train to proceed. This would prevent, at B, a driver 
going to C, being handed a tablet or staff for the B-A 
section, also, the signal at B being " off " when a tablet or 
staff was out at C for a train to go to B. 

\Yhere tablets and electrical train-staffs are exchanged at 
speed, such interlocking becomes the more necessary. 
This is done in America and is a guarantee to a driver that 
the staff or tablet is ready for him. 

If the arrangement went further and the distant signal 
could not be lowered unless the staff or tablet were on the 
catcher, it would be a greater improvement. The reason 
the Author has for suggesting the distant signal is that if the 
staff or tablet had to be on the catcher before a stop signal 
were lowered it would prevent a staff or tablet being handed 
to a train waiting at the signal-box. 

Fig.. 190. 

Whitaker's Exchanger for Train-Staffs and Tablets. 

Fig. 191. 



Fig. 192. 

Unlocking Outlying Sidings. 

There are several appliances whereby siding points on 
single lines can be worked from ground frames which are 
unlocked by, and interlocked with, the electric train-staff or 
tablet. Where these are provided no running signals nor a 
signal box are necessary for the protection of the connection, 
as the staff or tablet is its own protection. If, however, 
Permissive Working is in force signals are required. 

Fig-. 192. Tyer's Siding Lock. 

Messrs. Tyer & Co. have, since Mechanical Railway 
Signalling was published, brought out an arrangement 
whereby those points that are operated by " throw-over "in- 
stead of upright levers, are controlled, and which is illustrated 
by fig. 192. 

Above the lever a is a bolt b worked by the handle c 
working in a lock attached to the lever frame. In the lock 
is a recess protected by the lid d. The handle c is normally 
locked, so that the bolt cannot be withdrawn, nor the lever, 
of course, moved. 

On a tablet being inserted in the recess the tablet forces 
away the obstruction that holds handle c. The lid d being 
shut down, the handle can be turned and the bolt withdrawn 
so that the lever can be worked. 

Immediately the lever is moved the flat portion e on the 
boss prevents the bolt being shot again. The tablet also 
is held in by the fact that the other end of the bolt comes 
against the tail of the lid d so that it cannot be lifted up. 



Webb & Thompson's Electric Train-Staff. 

The electric train-staff designed by the late Mr. F. W. 
Webb and Mr. A. M. Thompson, of the L. and North 
Western R., illustrated by figs. 193 to 205, and manufactured 
by the Railway Signal Co. , appealed forcibly to mam- rail- 
way officials, because it retained the form of staff to which 
drivers were accustomed. It has deservedly become popular 
and is in use on long lengths of railways in all parts of the 

Fig. 193 illustrates the instrument and the staff. In the 
column is a deep slot which holds the staffs. To withdraw 
a staff it has to be raised into the drum-head at D, pass 

along the slot E and out at F. The staffs can easily be 
raised up to the slot E but cannot get further until certain 
discs inside the machine are freed electrically. The two 
springs, G 1 G 1 , are provided to hold a few staffs from going 
to the bottom of the standard, so that the signalman has 
not far to lift a staff. The mechanical pointer, H, is to re- 
mind the man whether a staff is out and whether for a train 
approaching or leaving him. The galvanometer I indicates 
when the current is flowing and a staff may be withdrawn. 
An electric switch J determines whether the current is to 
ring the bells or excite the staff coils, and K is the key by 
which the electric current is transmitted. The cross section 



Figs. 193-194-195. Webb and Thompson's Electric Train-Staff. 

Fig. 195- 



g Ti i f 


Fig. 196. 

Fig. 197. 

Fig. 198. 

Fig. 199. 

Fig. 200. 

Fig. 201. 

ffln - 

\i\ at 
' 1 

Fig. 202. 

Fig. 203. 
I )f tails of Webb and Thompson's Electric Train-Staff. 

ig. 204. 

of the column X X shows how the rings on the staff prevent 
it being withdrawn except at the aperture F in the drum- 
head. The rings L M N O on the staff are used for lifting 
pawls in the drum-head as described below, and P is a key 
for unlocking intermediate sidings. 

Fig. 194 shows an enlarged view of the drum-head with 
the front removed and a plan view of the drum-head with 
the top removed. Q is one of the five discs g, h, i, }, k, turn- 
ing on a common centre R and having four notches, S, cut 
at equi-distant points round their peripheries, equal in depth 
and width to the diameter of the staffs ; the discs are arranged 
as shown on fig. 196 and are shown in detail, figs. 197 to 
202. T is a pin upon which the mechanical locks or pawls 
are centred that secure the discs in position ; U is a lever also 
turning on the centre T and carrying the coils V. V. of an 
electro-magnet ; W. is a wedge-shaped lock attached to the 
lever W 1 , which also turns on the centre T, and this lever 
forms an armature for the electro-magnet V and is lifted 
should a current be passing through the coils (fig. 199) ; X 
is the tail piece of the lever U which carries the electro- 
magnet coils, fig. 194, and as this tail is foul of the slot K, it 
follows that when a staff is withdrawn or placed in the in- 
strument the coils V are raised with the lever U ; Y is an 
automatic commutator which switches the current alternately 

from the top to the bottom line wire, as the discs are turned 
a quarter of a circle, Y 1 Y 2 being projections on one of the 
discs (k, figs. 201-202), and which actuate the commutator 
J and also the key K ; Z is a side view of one of the five 
switches a, b, c, e, f. shown on the side elevation on fig. 194; 
a and b are fingers on the spindle J, fig. 204; c is actuated 
by q on the key spindle K, fig. 203 ; and e and f are moved 
by the commutator lever Y, fig. 202. 

The disc g, fig. 197, is a plain disc and has no pawl. The 
disc h, fig. 198, is similar but has one pawl, I, which turns 
on the centre T and which has a tail piece m. This slightly 
fouls the recess S so that when a staff is put in the instru- 
ment it causes the pawl I to rise and allow the disc to turn 
to the left, whilst, when a staff is withdrawn, the disc itself 
raises the pawl. This disc comes into play when a staff is 

The disc i, fig. 199, is the principal lock in the instru- 
ment. Its purpose is to prevent a staff being improperly ob- 
tained. It cannot turn to the right to allow a staff to be taken 
out owing to the lock W. \Vhen a staff is withdrawn it 
lifts the tail piecv X and this raises the lever L' carrying the 
magnet coils. Should a current be passing through the coils 
they attract the armature \\ -1 , and U and \V rise together, 
and so the lock W liberates the disc i and it is free to turn. 



The disc i can always turn to the left, and the disc h to the 
right. Should W 1 not fall back into its place after being 
freed, the lever P remains suspended and prevents the key K 
from being used for signalling until the matter has been put 
right, so that it follows that no second staff can be withdrawn 
without permission. 

Fig. 200 shows the mechanism which prevents staffs being 
withdrawn by illegitimate means. The disc / has two pawls 
o, centred at T, engaging in the top notch S ; the tail piece p 
of the first pawl is precisely similar to m in fig. 198, but the 
tail piece p l of the second pawl is shaped to prevent the disc 
; from being moved in any direction until the lock has been 
lifted clear by the staff raising the tail piece. This disc has 
also four shallow notches q which are cleared by the projec- 
tion r on the tail piece p l raising the pawl o clear of q as 
the staff passes. 

The disc fe, figs. 201-202, has two projecting cam pieces 
V 1 \~- for lifting the commutator Y, which is shown, fig. 202, 
in dotted lines, as lifted and actuating switches e f. Fig. 203 
is a side view showing key K actuating the switch c, the 
dotted lines indicating the reversed positions of the switch, 
and fig. 204 shows the two remaining switches a and b 
actuated by the cam J. 


I'ijj. 205. Webb and Thompson's Electric Train-Staff. 

Fig. 205 is a diagram of the connections. It is assumed 
that a train is to proceed from station B to station A and 
that the signalman at B has obtained permission to take a 
staff. He first reverses, by means of the pointer H, the 
position of the switch a, and joins up the local battery and 
switch a 1 from the bell contact to the staff contact, and 
thereby sends a current from the local battery b through the 
battery switch a to the local coils B of the magnet C. A then 
turns his pointer J, fig. 203, which reverses his switch h, fig. 
205, and this completes a circuit from his battery c through 
'* > / e > d, c, to line, and entering the instrument at B at c, 
passes through e, f, g, h, ', to /. The switch a 1 had already 
been reversed as stated above, so the circuit is through fe to 
the coil B and thence to /, m, n, and earth. The battery c at 
A is therefore in conjunction with the local battery, so that the 
latter excites the coil B and the former excites the coil C so 

that the magnet D is energised so that the lock \Y, fig. 199, 
is lifted and the disc / may be turned and a staff may be with- 
drawn. The withdrawal of the staff, seen in the description 
of fig. 202, reverses the position of the switches m, n, d, o 
in the instrument at B- If A were now to send another cur- 
rent it would again enter at c and pass through o. Mo the 
coils B in the reverse direction. This would be opposed to 
the current passing through C so no magnetism would be 
induced and the lock \Y could not be raised and therefore 
no staff withdrawn. 

If the staff withdrawn from B were restored to the in- 
strument there the disc fe, fig. 202, on being reversed would 
reverse the commutator lever Y and restore the switches e f 
(in n and d o, fig. 205), so that the instruments at A and B 
would synchronise. Similarly if the staff were taken through 
the section and put into the instrument at_A. 

" Permissive " Electric Train-Staff. 

On several Colonial and Indian railways, and also on the 
railways of the United States and South America, a permis- 
sive system is worked which precludes the use of the electric 
train-staffs already described, and this has led the manufac- 
turers to modify the instrument so that it can deal with per- 
missive working. In principle the instrument is the same 
as the \Yebb-Thompson, but only one staff is employed. This 
staff does not leave the station, but is withdrawn for shunting 
purposes or for exhibiting to a driver, to prove to him that 
he may accept a ticket. These tickets take the form of 
metal discs, and a considerable number are provided in each 
instrument. The working of the apparatus may be described 

as follows : 

A train is waiting at, say, station A to go to B- A signals 
to B, B sends an electric current to A which frees the staff, 
A withdraws the staff, and then rings to B for permission 
to withdraw a ticket, B turns an indicating dial which is 
provided on his instrument from o to I, and he is then able 
to send another electric current to A which frees the ticket. 
A withdraws the ticket and hands it to the driver of the train 
and shows him the staff as his authority to accept it. In 
like manner a second, third, or any number up to nine, tickets 
may be withdrawn by A- In each case B turns his indicating 
dial for each ticket before he can give the necessary permis- 
sion to A- The first train having arrived at B the ticket is 
placed in the instrument at B, and the insertion of the ticket 
automatically turns the dial back one number and so on 
until all the trains despatched from A have arrived. When 
the last ticket is placed in the instrument at B the indicator 
dial is turned back from I to o, and only when this is done 
can A give B power to withdraw a staff and ticket for a train 
to leave B. If a ticket be withdrawn by A and it is found 
necessary to cancel the running of the train for which it has 
been withdrawn to admit of a more important train to be 
forwarded from B, a case which frequently occurs with the 
" Permissive " system, arrangements are made for A to re- 
place the ticket in his instrument and to send an electric 
current to B which enables B to turn his indicating dial back 
from i to o. \Vhen this is done B is in a position to send 
the more important train forward. 

H 2 









by giins. wide. The length of the staff is lofins. and it 
weighs only 9^ oz. 

Fig. 206 shows the " M " type miniature instrument with 
a telephone attached. The telephone is coupled to the same 
line-wire as the staff instruments, so that only one line-wire 
is required for the instruments, bell communication and 

Though the general arrangements of the instrument re- 
main as before, the modifications noted below are improve- 
ments. The staffs are kept in four slots instead of in one, 

Fig\ 206. 

Miniature Electrical Train-Staff. 

The Railway Signal Co. have now introduced a new form 
of electrical train-staff. The internal construction of the staff 
instruments remain practically as before, the leading feature 
being the reduction of the size of the train-staff and a conse- 
quential decrease in the dimensions of the instruments, the 
total height of which are now only 2ft. 4ins. by gfins. broad 

Coil Lever 

Contact Rubber for operating 
the Local Battery Snitch by means 
of Special Lerer 

Fig. 2076. Staff lifted into Instrument Head, Coils not energised. 

Fig. 207A. Normal Position. Fig. 2oyC. Staff lifted into Instrument Head, Coils energised. 

Fig. 207. Miniature Train-Staff, Special I. ever for Locking Armature Lever until Coils are energised. 



Normal rosition of Automatic Lever. Of.-r Position of Automatic I.e-.-er. 

Fig- 208. Miniature Train-Staff, Lock on Automatic Lever. 

so that the same staffs are not used over and over again. 

A lock has been added on the armature lever (XV 1 in figs. 
194 and 199), which holds the lever down until the coils 
have been energised. This is illustrated in three positions, 
A, B and C, fig. 207. A special lever a has been provided 
which is pivotted at b, the tail c of which is lifted up when a 
staff d is raised to be withdrawn. On the armature lever e 
is a pin f engaged normally with a jaw g l on the special lever, 
as in position A, fig. 207. 

Should a staff be lifted for withdrawal before the coils h 
have been energised the special lever a is turned to the right, 
and the pin / simply passes from under one jaw g l to under 
the other jaw g-, as seen in position B, fig. 207. 

But if the coils be energised and a staff may be withdrawn, 
the armature lever e will be attracted bv the coils h imme- 

diately the staff turns the special lever a, and as its jaw g 1 
frees the pin / the lever e will leave / clear of the jaw g-, as in 
position C, fig. 207, so that in combination with the other 
discs and locks, already described, the staff is withdrawn. 
One object of this additional attachment is to guard against 
the discs being strained. There is also a short connection ; 
from the special lever to the switch k for switching in the 
local battery, which is done automatically instead of by the 
signalman with the pointer J in figs. 194 and 204. 

Another improvement is shown by fig. 208. The disc k 
in figs. 194 and 202 has had two locks added to it, one, a 1 , 
to prevent the lever being illegitimately lifted when in the 
normal position, and the other, a", to prevent the lever being 
improperly pulled down when in the " over " position. 

The illustration in fig. 206 should show a dial and in- 

Cut Out Snitch 

Cut Out Sr*;tcn 

I-HB ^ f -BottomTermtnal 'of 'Arrester 


Fig. 209. Miniature Train-Staff. Wiring for two Instruments working in Sympathy with each other. Without telephone. 



dicator on the left to advise the signalman when a staff is out, 
and, if so, whether it is for an up or a down train. 

Fig. 209 shows the wiring for a pair of instruments. 

The arrangements for cutting out a staff station at night or 
on Sunday are as follows : Assume that there are two sec- 
tions, A-B and B-C, and that B has t(J be closed at night. A 
separate pair of instruments, operated by an additional line- 
wire, is provided between A and C. In the signal-box at 
B there is a case with three drawers, in one of which is kept 
an A-C staff. This staff being out of the special instruments 
breaks them down and they cannot be worked. 

Assuming that the last train before B has to be closed 
is coming from A- The man at B will ask C for permission 
to withdraw a B-C staff in the usual way. This he will 
place in one of the vacant drawers, and on receipt of the 
A-B staff brought by the train he will place this staff in the 
remaining drawer. The presence of these two staffs will 
release the locked-up A-C staff wherewith the train will pro- 
ceed to C- Whilst the other two staffs are out, the A-B 
and B-C instruments are broken down and no staffs can be 
withdrawn. This efficiently switches one lot out and the 
other in. 

As the miniature train-staff is hardly strong enough to 
act as a key for unlocking sidings, as shown in fig. 210, the 

Main L,ne 

Fig. 210. Lock for Outlying Sidings without Key. 

new arrangement, illustrated by fig. 211, has been devised. 

At the outlying siding, fig. 210, a ground lever is pro- 
vided, and in the rodding from the lever to the points a 
staff-lock, fig. 211, is fixed. 

In the lock are four levers a, turning on c, the longer ends 


Fig. 2H. Miniature Train-Staff Lock. 

of which rest in the plunger coupled to the point rodding, 
and until these levers are raised clear the ground lever cannot 
be moved. In the box is also a drawer b which is pulled out 
when the points have to be moved, and the train-staff is 
placed in it as shown. When the drawer is pushed in the 
four wards on the staff come in contact with the short ends, 
and raise the long ends, of the four levers u and free the 
plunger. When the plunger is drawn along, the longer 
ends of the levers cannot drop owing to their removal from 
the slot in the plunger, and the short ends of the levers 
back-lock the drawer and retain possession of the train-staff 
until the ground lever is restored to its normal position. 

Electrical Tablet and Electric Train-Staff Apparatus for 
Non-Crossing Stations on Single IJnes of Railway. 

It is occasionally necessary to have a tablet or electrical 
train-staff exchange station at a place where trains cannot 
pass each other owing to want of accommodation. In such 
cases the possibility of trains being accepted from opposite 
directions must be avoided. 

The following are the alterations and modifications 
effected by the late Mr. F. T. Rollins in Tyer's No. 6 Tablet 
instrument (see fig. 166) to attain the object in view, which 
is to render it impossible for the signalman at an interme- 
diate non-crossing station to accept a train, and release a 
tablet, or train-staff, in both directions at the same time. 

Having accepted a train and released a tablet or staff 
from the signal-box on one side of him, it is rendered im- 
possible for the operator to accept another train or for a tablet 
to be issued from the other signal-box until the sections on 
both sides of him are again clear, although he may accept 
following trains, and thus have a train in both sections going 
in the same direction. 

In applying the arrangement to the Electric Train-Staff 
instrument, the necessary modifications will readily suggest 
themselves to those using that apparatus for working single 
lines of railway. 

Fig. 212 is a diagram of the circuit arrangements. 

At the intermediate non-crossing station B there is (in 
addition to the usual tablet apparatus with several of the 
connections altered and additional parts inserted) two polar- 
ised relays, 2. and 3. The relay 2 is in connection with the 
tablet instrument working to A, and relay 3 is in connection 
with the instrument working to station C- When the tablet 
instrument at B, working to station A, is plunged in order 
to allow A to get a tablet, a current from a local battery, 4, 
at B is, by means of independent contact 5 (13 is the corres- 
ponding contact for the B-C section) pressed together by 
the ordinary plunger, made to operate polarised relay 2, and 
by this means breaks down the main plunging battery 6, 
of the other tablet instrument working to C, so that a train 
cannot be accepted from the latter place now that one has 
been accepted from A. These polarised relays, 2 and 3, are 
also in the line circuit of the apparatus, preferably between 
the tablet relay coil and the earth, or metallic return ter- 
minal, so that immediately the commutator is turned to 
get a tablet at A, the next current sent from that place would 
actuate the polarised relay 2, and break down the main 



battery 6, at B, controlling the B-C section (10 is the cor- 
responding battery for the A-B section), even if the first 
plunge at B, giving permission for the commutator to be 
turned, had not already done so. And again, when the tablet 
is restored to the instrument either at A or B, and the ap- 
paratus reset, and the commutator turned to its normal posi- 
tion, a plunge from A will operate the said polarised relay, 
2, in the reverse direction, and thus rejoin the main trans- 
mitting battery, 6, of the apparatus working to C- To en- 
sure, however, that the apparatus shall not be completely 
restored, and thus a train be accepted from C whilst a train 
from A is standing at B (or shunting in adjacent sidings), a 
third polarised relay, 7 (which can be termed the governing 
relay), is provided at B, the coils of which are in circuit, 
one marked A" with an inner 8, 8 1 , and the other marked I" 
with an outer 9, g l rail contact or treadle on either side of, 
and at a suitable distance from, the signal-box at B- The 
outer rail contacts 9, g 1 on both sides, should be at a suffi- 
cient distance to be, at least, a train's length beyond the 

Station A. 

signals for trains going in that direction (and out of range 
of an}- shunting operations), so that any train passing over 
them must be on its way through the section. The two 
inner rail contacts 8, 8 1 should be nearer to the box, as it 
does not matter if, in shunting, or through the train stand- 
ing upon them, they are operated. The circuits are so 
arranged that, on a train passing over either of the outer 
rail contacts 9, g 1 , the relay is made (so far as the governing 
relay is concerned) to join up the main batteries for both 
tablet instruments, if not already joined up; and on passing 
over the inner rail contacts 8, 8 1 is operated in the reverse 
direction, and this breaks down the main plunging batteries 
of both instruments. Therefore, as the inner rail contacts 
are always operated last, on the arrival of a train, both 
main plunging batteries, 6 and 10, are broken down, and, 
although all tablets may be in the instruments, the apparatus 
cannot be completely restored as long as the train remains 
there. \Yhen, however, a train goes right away and over 
the outer rail contact g 1 with the tablet for that section, the 

Station C. 

f I 

i ; 

Fig. 211. Hollins' Arrangement of Electrical Tablet and Electric Train-Staff Apparatus for Non-Crossing Stations. 



main batteries, so far as the governing relay is concerned, 
are again joined up, and the signalmen may then clear up 
and completely reset the apparatus for the section in the 
rear, and, if required, another following train may then be 

There is also a further extra contact, n and 12, fitted 
in the tablet instruments at the non-crossing station B. 
When either A or C has given B permission to take a tablet, 
the lower disc of the non-crossing instrument shows " tablet 
out," and this disc, being in this position, closes a spring 
contact ii and 12, and joins up the negative pole of the main 
battery so that a negative current may be available at B to 
operate the upper disc of the tablet instrument, from which 
permission was given to obtain a tablet. 

Now, it is obvious when the main batteries are broken 
down at non-crossing station B as described, the signalman, 
unless other provision were made, would be deprived of any 
means of communicating with stations A and C, either to 
acknowledge a sound signal received, or to ask for a tablet 
for a train approaching him. The necessary sound signal 
communication is effected (if only one line wire is available) 
upon the wire used for the tablet circuit, by employing an 
induced vibratory alternating secondary electric current. 
This current is produced by means of an induction coil, with 
ordinary vibratory make and break 15, in the primary circuit 
16, and an electrical condenser 17, of suitable capacity, in- 
serted in the secondary circuit 18, which is arranged as a 
shunt from the line wire 20. A telephone receiver 21, capable 
of emitting a loud sound when operated by such a vibratory 
or alternating secondary current, is employed as a receiver. 

A commutator 19 is fixed in such a position in the tablet 
instrument, that when the ordinary tablet plunger is operated, 
a local battery 22, transmits a current through the primary- 
circuit 16, and make and break armature contact 15 of the 
induction coil 14, and thus induces a rapid alternating 
secondary current, in the secondary circuit 18. This, freely 
acting by induction through the condenser 17, and so to 
the line wire (and a similar condenser and other apparatus 
at the opposite end of the circuit) emits, through the said 
telephone receiver, a loud audible signal, which serves the 
purpose of the ordinary bell and gong signals usually em- 
ployed. The ordinary primary direct currents used for the 
tablet apparatus do not act with sufficient effect through the 
condensers to appreciably affect such telephone receivers ; 
and neither does the alternating secondary current, which 
acts freely through the condensers, affect the tablet 
instruments, and, therefore, there is no confusion in 
using either separately, or simultaneously, a primary 
and a secondary current upon the same line wire, 
one to operate the tablet apparatus proper and the 
other to give the sound signals. For this purpose an induc- 
tion coil 14 is used, and a telephone receiver 21, an ordinary 
carbon telephone transmitter 23 is added in the primary cir- 
cuit, and this also gives a telephonic communication between 
the two points. The auxiliary telephone receiver 24 is sus- 
pended from the switch lever 25, and so depresses the latter, 
as shown, into contact with contact point 26, which puts the 

loud telephone receiver 21, direct to earth. When the 
auxiliary receiver is taken off the switch hook 25, it joins 
up the telephone battery, the induction coil, and the micro- 
phone 23, by means of contact 27, which is insulated from 
the switch lever 25. In addition to this it also diverts the 
incoming secondary current from receiver 21, through the 
switch lever 25, and, instead of direct to earth, to the 
auxiliary receiver 24, and thus to earth E, through the 
secondary wire of the induction coil 14. 

Of course, by means of a second line wire, the ordinary 
bell and gong call signal may be retained, if desired, by 
merely having an extra independent battery contact, or com- 
mutator, to be operated by the ordinary plunger, and so 
transmit a battery current on the second line wire, for the 
bell and gong communication. To this may also be added, 
if desired, ordinary telephone communication. 

Hansel's Electric Train-Staff. 

The Webb & Thompson staff was first fixed on American 
railways by the Chicago, Milwaukee and St. Paul RR. in 
May, 1894, and the following is an extract from a paper 
read shortly afterwards before the Western Railroad Club, 
by Mr. C. A. Goodnow, then Assistant General Superinten- 
dent of the Chicago, Milwaukee and St. Paul RR., and late 
General Manager of the Chicago and Alton RR. : 

The lines of the Southern district of the Chicago, Milwaukee and 
St. Paul RR. cross the Mississippi river between Savanna, Illinois, and 
Sabula, la. The distance between these two stations is three miles, and 
there is one grade crossing, one draw-bridge, and one local station in 
ihe block. Over this track, which is single, the traffic of about 3,000 
miles of the St. Paul company's line passes. These lines extend directly 
to Kansas City, Omaha, Sioux City and Chamberlain on the west, and 
lo Chicago, Milwaukee and Racine on the east. During a larger part of 
the year the traffic is heavy (the bridge block being the neck of the 
bottle, so to speak) and will rarely fall below fifty trains per day at 
any time. 

The division yard is located at Savanna, on the west side of the 
Mississippi river, making it necessary for the trains of both divisions 
west of the river to use the bridge block, and, moving the traffic from 
so large a territory, it is to be expected that they will be irregular in 
number and that they will bunch during certain hours. The use of a 
time table showing the trains over the bridge block was aban- 
doned, because it was found impossible to so arrange it that it was a 
reasonably correct exhibit of the traffic. Nor was it possible to move the 
trains through the dispatchers of either division, as the work on their 
respective divisions would not permit the close attention 10 the bridge- 
block which the nature of the service demanded- For a time in the 
early history of the bridge this was done, but the work was finally put 
in the hands of the operators at each end of the block. It was found to 
be necessary to use some other than the ordinary dispatching systems. 
That was found to be too slow and cumbersome to meet the require- 
ments of the quick work necessary under the conditions constantly 
arising incident to unexpected delays, and to increase or decrease traffic. 
To meet the conditions described a train order by card system was 
adopted, which was in successful use for many years. It was virtually 
a staff system the card representing the staff but it lacked one 
element. It was impossible to interlock the cards. And as traffic in- 
creased, and the acceleration of trains became necessarv, it was apparent 
that the company would be compelled to either double track the bridge 
block or find some unobjeclionable way of handling the trains. Owing 
to the character of the country the construction of a second track would 
have been very expensive, and the selection of a satisfactory system 
for handling the traffic between these points became the subject of much 
thought and investigation. After a thorough examination and inquiry 
the Webb-Thompson electric staff system, largely in use on the London 
and North Western R., and in Australia, was adopted and placed in 
service in May, 1894, the first installation of the staff system in the 
United States, and probably in either of the Americas. 

Mr. Charles Hansel, then of the National Switch and 
Signal Co., designed an improvement on the staff whereby 
the weight to be handled was 4^ ozs. instead of 2f Ibs. and 
" Permissive " working was added. The series of rings on 
the staff were made removable, and when a train had to be 
sent under permissive working the signalman withdrew a 
staff and took the rings off and sent the train through the 



section with a ring as authority. The last train took the 
staff and any rings unused, and on arrival at the other end 
the rings were fixed on the staff and it was placed in the 
instrument. It will be remembered that the rings play an 
important part in the electrical arrangements, as they lift 
the pawls in the instrument so that if a staff were inserted 
without them the unlocking ctould not take place. 

A decided objection to this scheme is the possibility of a 
ring being lost or getting bent and not going on to the staff. 

In America they have a very good method of dividing 
the staff and giving one half to the engineman and the other 
to the brakesman, so guaranteeing that the whole of the 
train has gone through the section, as the staff must be 
whole before it can be placed in the instrument. It is also a 
certificate to the brakesman that his driver has authority to 
proceed. Where rings are used instead of a staff the driver 
has one ring and the brakesman the other. 

Releasing Starting Signal by Electrical Train-Staff or Tablet. 
Mr. Hansel also provided an electric slot on the starting 
signal, whereby the staff or ring must be in position to 
be lifted by the engineman before the signal can be lowered, 
and also that directly the staff or ring is lifted by the engine- 
man the signal goes to danger. 

Ring of Electrical Staff 
in Position. 

Ring of Electrical Staff 
taken att'ay by Locomotive. 
Hansel's Train-Staff. Fig. 214. 

The engineman thus knows when he sees the signal 
" off " that the staff or ring is there for him. 

This is illustrated by fig. 213. The signal is on the left 
(with the arm pointing to the right in accordance with 
American practice) ; in the centre is the staff crane with a 
steel ring supported by two arms. The arms are electrically 
connected to the slot and the circuit is completed when the 
steel ring is in position. On the steel ring is carried the 
ring off the staff, and on the right is a catcher on the loco- 
motive, which, when it comes along, gets hold of the steel 
ring and carries it away with it. This causes the two arms 
to separate as seen in fig. 214, and the electrical circuit 
being broken the slot goes on and the signal to danger. 

Union Switch and Signal Co.'s A'eu' Train-Staff. 
As licensees for the manufacture of the Webb and 
Thompson Electrical Train-Staff, the Union Switch and 

Signal Co. modified it to suit American conditions, and their 
new staff is meeting with much favour. 

The machine, which is illustrated by fig. 215, is 2ft. gjins. 
high (5ft. 3fins. with pedestal), ift. 2^ins. wide at the widest 
part and ift. 3ins. wide at the base of the pedestal. 
Normally the instruments, with all the staffs in, are as 
shown (except that the lower door is kept closed) by fig. 
215. As in the original staff instrument, only one staff can 
be withdrawn at a time out of one of a pair of instruments 
applicable to the same section. Let it be assumed that a 
section extends from signal-box A to signal-box B- There 
would, as usual, be an instrument at A in circuit with one at 
B. If a train has to travel from A to B, the man at A would 
exchange signals, as per code, on bell plunger a, which would 
ring a bell in the instrument at B- The man there acknow- 
ledges the signal on his bell-key b, and then by keeping it 
depressed a releasing current is sent to A- This is indicated 
by the deflection of the lower part c of the indicating needle 
to the right. The man at A then turns the preliminary 
spindle handle d to the right as far as it will go and then 
allows it to return automatically to its normal position. 
This action frees a staff and the disc c in the indicator 
changes from red to white. A now turns the outer guard 
/ to the left so that the slot f 1 is lineable with the end of slot 
g, up which he passes a staff, revolves the outer guard 
through half a circle, using the staff as a handle, and then 
takes the staff out at /i. 

This reverses the polarity of the current, and causes 
" staff-out " to be shown by the upper indicating needle /. 
A then gives one more beat on his bell a, which indicates 
to B that the operation is completed, and causes his upper 
indicator to show ' ' staff-out ' ' also. 

On arrival at B the outer guard is turned to the central 
position, the staff is inserted through the slot of the outer 
guard into the slot of the machine, and the staff is placed 
amongst the other staffs on the right. 

As all American lines have at times to be operated on the 
Permissive " system, allowing a second and other trains to 
follow before the first has passed out of the section, a " Per- 
missive " staff has been provided, and this is carried in a 
separate attachment on the right of the instrument as seen 
in fig. 216. It consists of a steel rod having eleven remov- 
abie rings, and is kept in the box a. When "Permissive" work- 
ing is resorted to a staff is withdrawn from the " Absolute " 
instrument in the usual way and inserted in the hole b. It 
is forced along the slot c, and in so doing a lock which holds 
the " Permissive " staff is removed, but the original staff 
is locked in. Each train travelling on the " Permissive " 
carries a ring, and the last train takes the " Permissive " 
staff with the balance of the rings. Those that have been 
used are placed on the steel rod at B, and the whole inserted 
in a recess in a similar attachment at B. There are certain 
wards in the recess which fit the rings, so that unless the 
" Permissive " staff is whole the operation cannot be com- 

When the '"Absolute" staff was placed in the "Permissive" 
attachment at A it allowed for an " Absolute " staff to be 



withdrawn at B and placed in the " Permissive " attachment 
there, but as there is only one " Permissive " staff for a 
section no evil result can arise. The reason for this apparent 
irregularity is to unlock the box to receive the " Permissive " 
staff, and until the "Permissive" staff, complete with its rings, 
is inserted the " Absolute " staff cannot be withdrawn at B, 
nor that at A, and until this is done, and each staff returned 
to its instrument, the circuits are not complete. 

The internal mechanism of both " Absolute " and " Per- 
missive " instruments is illustrated by fig". 217. When a staff 
is withdrawn the drum a is engaged, which, when revolved, 
reverses the polarity of the operating current. When a staff is 
inserted it engages the drum b and causes the machines to 

In the " Permissive " machine the hole in which the "Abso- 
lute " staff is inserted is at c, and the wards d make and break 
electrical contacts in e. The " Permissive " staff is in / and 
held by locks operated by the rod g. 

The " Permissive " staff with its rings will be noticed 
in fig. 218. Here the "Permissive" attachment is on the left, 
whilst on the right is another attachment which is provided 
where heavy gradients necessitate the use of assistant 
engines in the rear to push the train up the bank and then 

return without going through the section, as is done outside 
Oban on the Callender and Oban R. 

To obtain a " Pusher " staff an " Absolute " staff must 
be inserted in the hole a, fig. 219, and passed along the 
slot to b. This takes out a lock, makes and breaks electrical 
connections and frees the "Pusher" staff c. The "Absolute" 
staff may now be withdrawn as well as the " Pusher " staff, 
but in such operations two conditions are essential an ' Abso- 
lute " staff is necessary to obtain a " Pusher " and the with- 
drawal of a "Pusher" locks up the whole instrument, and 
neither an " Absolute " nor a " Permissive " staff can be 
withdrawn until the " Pusher " staff has again been re- 

An advantage of the new staff instrument is the small 
space it occupies, and the electrical repairmen find an advan- 
tage in being able to get to the connections without moving 
the machine, as the upper part turns on a centre as seen 
in fig. 220. 

For controlling intermediate sidings in a section a lock 
is provided on the lever as seen in fig. 221. By inserting 
the staff at a and turning it to b a lock is taken out of the 
catch rod c, and the lifting of the catch rod locks the staff 
in so that it cannot be withdrawn until the lever is restored 
to its normal position. The recess in 
the quadrant for the catch rod is shal- 
lower in the " over " position. It is there- 
fore ensured that no siding can be 
used unless the train has a staff, that 
the staff cannot be carried forward until the 

Fig. 215. 

High-Speed Electric Train-Staff. 
Instrument with all Staffs in. 

Fig. 216. 
Staff Instrument with " Permissive 


Fig. 217. 

"Absolute " and " Permissive " Staff Instrument. 
Internal Mechanism. 



Figs. 218 and 219. 
Staff Instrument with " Permissive " and " Pusher " Attachments. 

lever has been restored, and, if worked with an "Absolute " 
staff, that no second train can approach whilst the siding 
is being used. 

Means are also provided for unlocking certain levers 
in an interlocking machine by means of the train-staff where 

it is advisable that such levers should not be used unless a 
staff be out. 

Where it is necessary to use an intermediate siding as a 
passing place it can be arranged for the shunted train when 
inside to insert its staff into an instrument and, turning the 
same, to lock the staff in, and in so doing to synchronize the 
instruments of the section. \Yhen the more important trains 
have passed the men at each end, being mutually agreed, 
can release the staff, but in so doing they reverse the polarity 
of their own instruments, which can only be restored by the 
arrival of the released staff. 

Any "Absolute " or " Permissive " staff withdrawn in 
error or for a train that has not to travel through the section 
can be restored to its original instrument, but when one or 
more trains have passed into the section on the "Permissive" 
system that " Permissive " staff must follow. 

The Author suggests that it be a rule where "Permissive" 
working is in operation that when a ring is handed to an 
engine driver he should be shown the " Permissive " staff. 
This would guard against trains being sent with a ring im- 
properly retained after the staff had gone. 

The staffs for adjacent sections are differently shaped to 
prevent their getting into a wrong instrument. 

There appear to be some objections to this "Permissive" 
form of working. For instance, " Permissive " working can 
only commence at that end where there is the " Permissive " 
staff. Also the man who has the " Permissive " staff mav 


Fig. 220. Shows Accessibility of Mechanism. 

Fig. 221. Siding Lock for L'nion S. and S. Co.'s New Train-Staff. 



send any number of trains regardless of whether the man at 
the far end can accommodate them. Further, there is the 
objection, already noticed, of using rings as tokens, as they 
can be so easily lost, and if bent cannot be replaced on the 
staff. Finally, when trains have to be sent from the other 
end and the " Permissive " staff is not there one more train 
must be sent to the other end to take the staff. 

The " Permissive " tablet system would seem to be 
better, as " Permissive " working can be commenced from 
either end ; the man at the far end must accept each train ; 
each train has the same token a tablet ; the flow from one 
end can be stopped and commence in the opposite direction 
at any time ; and, lastly, the number of trains to follow one 
another is not limited as with the rings on a staff. 

- 423- Facing Points (see p. 263). 

Fig. 481. Signal Post and Ground Disc Signal (see p. 265). 

The first installation of the " Crewe " System of " All-Electric Pow/r-workod Signals and Points laid down outside the Locomotive 

Department Offices at Crewe, L. and N.W.R. 




Automatic Signah for Single Lines. 

IT has already been said that in America automatic sig- 
nals are used for protecting single lines, and the extent to 
which this is being done may be judged by the record (see 
page 79) of the Harriman and the Queen and Crescent lines. 

The arrangements vary according as to whether the sec- 
tion is long or short, and this is determined by the volume 
of traffic. On the Queen and Crescent the sections average 
ii miles in length. No hard and fast rule exists as to such 
signalling, and in some cases, for instance, distant signals 
are not provided. The points at crossing places are not 
worked as a rule from signal boxes, but from the ground, 
and the automatic signals are interlaced through the points, 
so that when a train lowers a signal the points are held fast, 
also when the points are lying for the opposite direction to 
that in which the train has to travel, the signals are kept at 

Such a system as that illustrated by fig. 222 might, in 
the Author's opinion, be adopted on British lines. The sig- 

"5 *6 

Fig. 222. Automatic Sicrnals for Single Lines. 


nals being automatic would not have to be worked from a 
locking-frame, and all that would be necessary would be two 
levers at the single line junctions one to work the points, 
and the other the facing-point lock. 

Three features would be essential. First, there must be 
the "Track Circuit" throughout; second, that the signals must 
be worked on the "normal-danger" system, so that the road is 
always regarded as blocked unless a train be coming, and 
the line be clear, and to avoid signals for the opposite direc- 
tion, and therefore conflicting, being off together, and third, 
that in order to render still more remote the possibility of 
trains departing from A towards B, and B towards A simul- 
taneously, the starting signal at A (signal i) should be elec- 
trically controlled from B, and signal 6 at B controlled from 
A- These would be electrically controlled, and could be 
worked by a switch in the station-master's office, and the 

switch should be back-locked by the "Track Circuit," so that 
the permission having been accepted, and a train started, 
it could not be withdrawn. This would be a further safe- 
guard beyond the fact that any vehicle standing past signal 
i would lock signal 6. 

The electrical connections to the signals would pass 
through the points, so that the points must be lying right, 
or the signals will not come off. For instance, No. 2 home 
would not be lowered unless the points near it were lying 
normal for the straight road, and these would have to be 
over before signal i would come off. 

Being on the " normal-danger " method the signals are 
lowered by approaching trains, No. i being lowered by an 
action set up by the train passing over a relay fixed some 
distance back, which, if the line be clear to signal 5, would 
lower signal i. Signal 5, followed by distant-signal 4, would 
be low-ered by an approaching train, providing the section 
ahead were clear, and the points set right. If the single line 
points near 5 were lying for the wrong direction, signal 5 
and consequently distant 4 would remain at danger, and the 
driver would pull up. 

For Indian, Colonial, and other lines where the signalling 
arrangements need not be so complete, the same method, 
modified to suit, could be adopted. The Author's ideas in 
reference to this are shown by fig. 223. It is assumed that 
the points are not coupled to a ground frame, but are 
weighted to lie for the straight road, so that trains running 
through them in a trailing direction simply burst the points 
open and they fall back to normal automatically. Not being 
worked in the usual way, the signals do not detect them. 
Nor is it proposed to control the signals from the distant 
station. As an alternative safeguard, it is proposed to pro- 
vide advance signals on the home signal posts so that signal 
9 cannot fall if a train is on the line on the A side of signal 9 
whether the train is going to A or coming to B- 

Winter's Single Line Working. 

Siemens Bros, and Co., Ltd., are the manufacturers of 
the system of single line working used in many parts of 
India, also in Australia and the Argentine, and which was 

Fig. 223. Automatic Signals for Single Lines. 




Fig. 224. Winter's Single Line Block 

designed by the late Mr. G. K. Winter, of the Madras R. 

Here there are a pair of block instruments with the signal 
for entering the section electrically controlled by the block 
instrument, so that it cannot be lowered for a train to enter 
the section unless permission has been given on the block 
instrument by the man at the other end of the section. It is 
also arranged that the train itself throws the signal to danger 
and therefore it cannot remain off for a second train. 

The instruments are shown by fig. 224, and indicate 
whether permission has been given for a train to come or 
permission received for a train to be sent. 
In fig. 225 are given the electrical details. 
The method of working is, that if a train is ready to go 
from A to B, the signalman at A commences by pressing his 
plunger P asking " is-line-clear?" If B can accept the train 
he will turn the switch handle at the side from " off " to 
" on," and will reply signifying " line-is-clear." By turning 
the switch handle ( x fig. 225) B reverses his line battery so 
that the signal thus given by him is in the reverse direction 
to signals given with the switch handle at " off." The action 
of this signal rings the bell at A and also reverses the 
position of the polarized tongue of a relay within the instru- 
ment. This relay controls the direction of a local current 
acting on the coils of the electro-magnet of the " train- 
going-to " indicator, but it cannot send the current through 
these coils. 

When A receives the signal indicating that the line is 
clear he presses the button V. This brings a local battery 
into play which sends a current through the starting signal, 
but this could not happen unless B had first sent the " line- 
is-clear " signal and A had closed the local circuit by pressing 
V. When he presses the button he keeps it pressed while he 
gives the acknowledgment on his plunger P. The result of 
this is to close the circuit of the local battery through the 

Fig. 225. Details of Winter's Single Line Block Instrument. 

coils of the indicator at A making it show " on line," and to 
disconnect the local battery from the signal control so that it 
cannot be unlocked a second time. Also it allows the cor- 
rect acknowledgment to be given on bell at B and his 
" train-coming-from" indicator to show " on-line." 

The control on the signal is fixed on the spindle of the 
arm and the arrangements are shown by fig. 226. 

The axle a is the spindle, the arm b on which is held by 
the catch c, which is on the centre d, and working on the 
same centre is an arm e. When the electro-magnet k is 
energised by the " line-is-clear " current already referred to, 
the armature i is attracted, which frees another arm g from 
the clutch h, and this allows the weight / to fall on to the 
arm e, forcing it down, and so taking the catch c from the 
arm b, and freeing the signal. When the signalman pulls off 
the signal the axle a revolves, and this causes the stud m to 

Fig. 226. Control on Signal, Winter's System. 



press against the stud u on the axle of the arm g and 
weighted arm /, so that the latter is turned to the right, and 
the catch h again secures the arm g, and all is in order when 
the signal is put to danger. 

In the event of the signal not being put back by the sig- 
nalman, it is thrown up automatically by the train. This is 
illustrated by fig. 227. The signal is not coupled directly to 
the usual balance lever L, but the rod R is raised when the 
lever / is moved upwards. When the signalman pulls over 
his lever to lower the signal, the lever L is raised and this 

between the teeth of a rotating tumbler in the interior 
mechanism of the machine. To turn the key round the 
tumbler has to be also turned, but its movement is con- 
trolled by electric locks actuated by currents from the next 
box. After exchanging the necessary bell signals the signal- 
man at K gives B permission to withdraw a key. The grant- 
ing of this permission frees the rotating tumbler and allows 
it to revolve, and so allowing the key to be turned and with- 
drawn. After a key is withdrawn the tumbler is again 
locked, so that another key cannot be withdrawn without 
further permission, and this permission cannot be given until 
tht key has been put into the instrument at the other end, or 


Fig. 228. Theobald s Kev. 

Fig. 227. Automatic Replacement of Signal, Winter's System, 
raises the cam c, which turns the lever I upwards, and takes 
with it the rod R. \Yhen the rod R goes up a cam k drops 
under the rod, and holds the signal off, even when the lever 
L is put back. The cam k is weighted, and it takes along 
with it the rod r, which raises the trigger / centred at P. 

When the train passes, the trigger / is depressed, the 
arm t turned, and the cam k drawn from under the rod 
R and the signal goes automatically to danger. 

Theobald's Electrical Key. 

Mr. Theobald, chief telegraph inspector, Madras R., in- 
vented a method of working single lines, the instruments for 
which are made by Orr and Sons, of Madras. Primarily 
intended for single line working, it is also capable of appli- 
cation to block purposes on double lines. 

This method takes the form of keys which are kept in a 
pair of instruments one at one end of the section and the 
other of the pair at the other end. The instruments are 
i Sins, long by ijins. high and 5ms. wide, the keys being 
Sins. long. The keys for the different sections of line vary 
in shape so that they cannot be placed in the wrong instru- 
ment. The instruments are worked with one wire. 

Fig. 228 illustrates the key used, and 229 is a side view of 
the machine, and fig. 230 is a front view. 

The keys lie in the slot M, and it will be noticed that they 
travel down the slot into the aperture M 1 . There they have 
to be turned, as the projection fc on the key in travelling 
down the slot is at the back of the key, and the key has to 
be turned so that the projection comes out first. \\ hen the 
key is in the aperture M 1 , the projection fe finds its way 

Fig. 220. Theobald's Key. tig- 230. 

it may be restored to the machine it was taken from, When 
a key is restored it cannot be put direct into the slot M but 
has to be let down into the aperture M 1 and turned, and in 
doing this the projection fe enters again between the teeth of 
the revolving tumbler, and turns it in the opposite direction, 
so restoring communication. 

The system also provides for " Permissive " working, 
and for a banking engine to assist with a train over part of a 

\calc 's Tokens. 

A system very much used in India was invented by Mr. 
]. E. Xeale, telegraph superintendent, Great Indian Penin- 
sula R. 

The tokens are balls, the issue of which is controlled by 
the instrument illustrated by fig. 231, a pair of which are 
provided for each section. 

When a train is to proceed from A to B the man at A 
presses in his bell plunger a which rings the bell b at B and 
unlocks the handle r at B, allowing it to be turned to the 
left. The man at B presses in his plunger, which allows A 
to turn his handle c to the right, when a ball falls out at d 
and is handed to the driver. On arrival at B the ball is in- 
serted in the drawer e and it travels along the zig-zag path 
as shown. In so doing it comes in contact with the vertical 
sliding shutter /, which, by means of the double relay ,<r, 
completes a local circuit and energises magnet /i', which lifts 
the lock / from the handle c and allows it to be turned to the 
vertical position. The plunger a being pressed in energises 
the magnet h- of the instrument at A so that the lever c there 




Vig~ 2$i. Instrument for Nellie's Tok 

can be returned to vertical and the instruments are again 
normal. A screen is provided at k to show what tokens are 
in and indicators at /' I- to show which way the handle c is 

//</>/>( '.v l\cy Instruments. 

C'a|)t. Hepper, of the Indian Railway Hoard, has designed 
the instrument illustrated by fig. 232 for regulating the issue 
of keys for working the loop points at the entrance to 
stations and other sidings. Ordinarily these are controlled 
from the station-master's ollice, but considerable time is 
taken up in getting the keys to and from the ollice and the 

Hv Capt. Hcppcr's arrangement there is a pair of instru- 
ments for each ground frame, one instrument in the ollice 
and the other at the points. 

When the points have to be worked a key has to be with- 
drawn from the instrument at the ground frame. The neces- 
sary bell signals are exchanged, and then the station-master, 
by turning the key in his instrument to the right, sends a 
current to the further instrument which allows the man there 
to turn his key to the left and withdraw it to put in the lock 
of the level" and to work the points. 

In the illustration the instrument at the ground frame is 
shown with the key withdrawn. The lock ti has a projection 
h that normally rests on the detent c at the end of the exten- 
sion of an arm-iture lever </ \\hich is attracted by the 
magnet c. 

The magnet c cannot, however, release the lock a unless 
the key be turned, as a pin / has to work in the slot cut in 

a lever g and this lever will only move sufficiently for this by 
the action of a shallow ward on the key, as in an ordinary 
lever lock. When a key may be withdrawn the magnet e is 
energised, causing the projection b to be freed, and on the 
pointsman turning the key to the left he brings the lock a 
down and takes out the key. At the same time the insulated 
pin /< is withdrawn from contact with the springs j' 1 j-. 

In order to give permission for the key to be withdrawn 
from the distant instrument the station-master turned his 
key to the right, which brought down the switch k which 
broke contact with /' /'-' and made contact with /' P. 

A spring in is provided to assist in bringing back the 
switch and a tappet lock n fitting between the lock and the 
switch so that only one can be used at one time. 

Kig. 233 is a diagram showing how these instruments are 
used in India, and for which the Author is indebted to Saxby 
and Farmer, Ltd., who are the makers of them. 









Sykcs' System. 

Mr. Sykes invented a method of working single lines 
\\itliout staff or tablet upon the principle of controlling the 
signals by the block instruments. 

When a train has to travel over a section of single line 
horn A to B the signalman at A asks B for permission, 
and the granting of this permission locks up the instrument 
at B and frees the instrument at A- Working the instrument 
at A unlocks the starting signal at A, and that back-locks the 
instrument there and it remains locked along with the instru- 
ment at B until the train has arrived at B, where the train 
pasM-s oxer an electrical treadle, which frees the instrument 
at B and allows the signalman there to give the " train- 
ptisscil-oiii-of-scciioii " signal to A so that the signalman at 
A can restore his instrument ready for permission to be given 
for a train to come from either A or B. The starting signals 
for entering the single line sec-lion are provided with auto- 
matic replaccrs, which necessitate the signal being put to 
danger for the first train before it can be pulled off for a 
second one. The lever in the locking frame cannot be worked 
until the block instruments arc properly set. 

Fig. .134 is a diagram of the signals for a single line sec- 
tion. Signal I is the starting-signal at A, which cannot be 
lowered until B has given permission for the train to 
approach him. Signal 3 is the home' signal at B. Klectrical 
treadle .' throws signal i to danger and treadle 4 reverses 
signal 3 and allows B to give the " train-passed-out-of-sec- 
ii,i " signal to A. Signals (i, 8, and treadles 5, 7, are for 
working trains in the opposite direction. 

In tig. J^s de-tails of the instrument used are illustrated. 
Tin- slide i is normally in the midway position and is drawn 

forward when a train has to be sent, and pushed to the back 
position when a train has to be received. In the case under 
notice the slide i in the instrument at A is held by the lock 
37, and when B sends a current the magnet 19 is energised 
and the lock 37 withdrawn. This allows the slide i to be 
drawn forward, so that the hole 3 comes under the hole 38 
and above the rod .24, which is coupled to the starting signal 
lever in the locking frame, which lever can now be pulled 
over and the signal lowered. 

\\ hen the slide is drawn forward the magnet 19 becomes 
de-energised and the lock 37 falls behind projection 44 on 
the slide and so holds the slide in the forward position. Re- 
placing the starting signal lever does not free the slide, as 
the lock 37 can only be taken out by the energising of the 
electro-magnet 19, which can only happen on a current being 
sent from B when the instrument there is put to normal. 
This cannot be done until the train has arrived at B and 
passed over treadle 4, as the corresponding slide at B, in 
order to accept a train, had to be pushed back so as to place 
in line the electrical contacts seen in the illustration. When 
B did this his slide was locked in the back position, in the 
same way as when A accepted the permission to send a train 
his slide was locked in the forward position. Hoth slides are 
then locked, but when the train passed over the treadle 4 tin- 
lock at B is taken out of the slide, which can then be drawn 
forward and a current can then pass to A, and entering the 
magnet 19 the lock is taken out and the slide can be 

There is another lock, 43, working in slot fi. This pre- 
vents a signalman accepting a train (i.e., pushing the slide 
from the midway to back position) unless the previous train 










had passed into the next section. This would be done at B 
by a treadle fixed at the other end of the station in a similar 
position to treadle 2 at A. The lock 43 is normally in the slot 
6, which is elongated so that the slide can be drawn forward 

/a r 6 

Fig. 234. Diagram of Signals, Sykes' Single Line System. 

when the lock 43 is in. When the train goes into the next 
section and the magnet 18 is energised, the lock 43 is raised 
and the clicks, 430, 43??, fall in when the armature is 
attracted. On the magnet becoming de-energised (the attrac- 
tion being only a momentary one) the armature rests on the 
clicks and the lock 43 out of the slot 6 and above the slide i. 
When the slide is pushed in, the thick part of the slide forces 
out the clicks so that the armature falls and the lock 43 rests 
on the top of the slide ready to enter the slot 6, when the 
slide is again put to the midway position. 

Details of the lock 43 are given separately in fig. 235, 
also particulars of the working of the indicator for showing 
the signals that have passed. In the face of the instrument 
is an opening, behind which is an indicator with " line- 
clear," " blocked," and " train-on-line." This is coupled by 
the rod 65 to a crank, one arm 63 of which binds the slide i. 
When the slide is pushed in to accept a train, the arm 63 is 
forced to the right by the bevel 2 on the slide, and this brings 
down the rod 65 and shows " line-clear." When the slide is 
drawn forward to send a train, the bevel 2 allows the arm 63 
to travel to the left, and this raises the rod 65 and shows 
" train-on-line." 

The system is in use between Clyde Junction and Gorbals 
for the up and down relief siding, fig. 567, and for the up 
and down carriage line at Victoria. The instrument is shown 
at the extreme right of fig. 566. It is largely used in Russia. 

Fig.. 235. Details of Sykes' Instrument for Single Lines. 

.S / ,V G L E L 1 A E S , SY K E S ' SY S T E .A/ . 


Fig. 236. Sykes' System, 
Cairn Yallev R-. G. and S.VV.R 

1 he Author has examined the working, which has since been 
further improved, and it seems to provide for all contin- 
gencies. The weakest point lies in the cancellation of trains, 
and here .it is provided that the station-masters at each end 
with a special key shall go to the treadles at their respective 
ends and by a special combined and simultaneous action they 
can make the same movements as would be made by a train. 
The Author, however, thinks that in combination with a 

' Track-Circuit " the system would be as near perfection as 
could be designed. 

It should, however, be remembered that when a driver 
has a staff or tablet he has some tangible proof that he is in 
possession of the section. In favour of the system it may be 
claimed that the driver has not two indicators to look for a 
signal and a staff or tablet. Time is also saved, as when two 
trains cross there is no need for the staff or tablet received 
from the arriving train to be put into the instrument before 
one can be taken out for the departing train. 

The Cairn Valley Light R. of the Glasgow and South 
\\ estern Co. has been signalled by the Sykes Interlocking 
Co. by an arrangement similar to that in fig. 236. 

YVhen a train has to be sent from A to B the man at A 
pushes in his bell plunger and sends a current from battery 
a" through line i to the bell at B and advising the man there 
that a train has to be sent. The signalman at B, if he can 
accept the train, pushes in his slide b 3 , which pushes the rod 
b 4 to the cheese cutter lock on the lever working the points 
so that they are held and the spring-pressed latch b 5 enters 
the notch b 6 and the switch b 1 enters the notch in the slide 
and at the same time breaks the circuit from battery b- and 
makes contact with b 8 . The man at B now presses his bell 
plunger for, say, 3 seconds, and this sends a negative cur- 
rent from battery b 9 over line i, through the indicator a 10 at 

A, through coil rt n to earth, so thnt the coil a 11 is ener- 
gised and its tongue a K rests against contact a 13 . A current 
thereon passes from battery 9 through the coils of the minia- 
ture semaphore to contact a 14 , bridge switch a 17 and contact 
a 18 to magnet a 19 , energising it and attracting the latch a 5 so 
that the slide a 3 can be pulled out. This locks the points at A 
in the same way as those at B, and at the same time switch 
u 20 is turned and joins up a- 1 and a~. 

When the slide b 3 was pushed in the switch b 20 made con- 
tact between b 21 and b 23 and consequently current of battery 
b 2 * now flows from earth through the magnet of signal B H 
so that it is lowered, through switch b 25 to battery b 24 , and 
thence through contacts b 53 b 20 b 21 to switches b 2 b 27 and 
line 2 to switches a 27 a 26 , contacts <J 21 a 22 , and thence to earth 
through the magnet of the signal A S for leaving A and 
lowering the same. 

The train may now leave, and in so doing it depresses 
the treadle a 28 , which joins up the battery a 29 to the winding 
of the " sticking " relay a 30 to earth so that the polarity of 
the relay is reversed and it makes contact with a 31 and conse- 
quently earthing the current of b 24 through the magnet of 
signal A S, so that the latter is put to danger behind the 
train. YVhen the train arrives at B it depresses the treadle 
b 28 and a circuit is then made through contact b 15 and bridge 
piece b 17 (the latter having been put into contact with b 15 
when the slide was pushed in) from battery b 29 so that the 
magnet b 19 is energised and the latch b 5 taken out of the slot 
b 6 and also for the slide to be returned to the normal mid- 
position and the latch b' to break contact with b 8 and join up 
battery b 2 . 

The signalman at B now presses his plunger for, say, 3 
seconds, so sending a positive current from battery b 2 over 



line i, through indicator a 10 , to earth through the relay wind- 
ing a 11 . This, being a positive current, moves the relay 
tongue and it makes contact with a 31 , with the result that a 
negative current is sent from battery 9 through contact a 16 
(upon which bridge piece 17 came to rest when the slide was 
pushed in) to the magnet a 19 , energising it and attracting the 
latch a 5 so that the slide can be put to the normal mid- 

Arrangements have been made whereby the concurrent 
action of both signalmen will unlock themselves in case a 
train has been signalled that does not subsequently proceed. 
The switches a- 6 a- 7 ?> 26 b 27 and the plungers i 2 are provided 
for this purpose. 

The indicator 10 shows when normal " line-blocked. " 
When the slide is out it shows " trahi-to-B " and when the 
slide is pushed in it indicates " train-from-B-" Similar sig- 
nals as to A are given on the indicator b w . 

In case a level crossing exists between two stations a 
current off line i causes an alarm bell to ring at the crossing 
and operates an indicator. Automatic signals are provided 
on each side of the crossing which work with the gates, i.e., 
they " clear " when the gates are across the roadway and 
open for the railway. 

Illinois Central Block ami oilier Systems. 
On the Illinois Central, Chicago and Eastern Illinois and 
other American railroads a serious attempt has been made to 
control single lines by means of a form of " Lock-and- 
Block " or "Manual-Control." 

No train can pass the signal at A for going towards B 
without the requisite block signals have been passed between 
the signalmen concerned and the man at B has taken off the 
lock on the signal at A. 

That the signal is put to danger for the preceding train is 
guaranteed by instructing the engineman to see the signal 
lowered. As said above, this is a serious attempt to deal 
with the question and deal with it economically, but this 
question of guarding against the signal remaining " off " 
reveals one of many objections. The Author has been told 
that American conditions forbid trains being delayed, and 
yet here, to meet the requirements of the system, the home 
signal must be kept at danger until the engineman sights it. 
Then, again, there is the possibility of trains breaking loose 
and the line being cleared by the first portion. This objec- 
tion is somewhat discounted by the fact that all trains in 
America are equipped with a continuous brake, which will, 
or should, pull up the train in case it became divided. 

Fig. 2360 is a view of the Illinois Central block instru- 
ment, which is manufactured by the General Railway Signal 
Co., and fig. 2366 is a view of the internal mechanism. 

A pair of these instruments are, as usual, provided for a 
section, one at A and one at B- When A desires to send a 
train to B the former exchanges the usual bell signals, and 
then B, if he can accept the train and if his signal for the 
opposite direction be at danger, turns the lever a, which 
joins up his battery and a current flows through the contact 
on the opposing signal and its lever, guaranteeing that these 

Fig. 236(1. " Illinois Central " Block Instrument. 

Fig. 2366. Mechanism of " Illinois Central " Block Instrument. 



are normal, through indicator b showing that he has given 
permission for a train to approach, to A, where it enters a 
corresponding instrument, energises the coils c so that the 
indicator d is raised, notifying to the man at A that he is 
unlocked and energising magnet e whereby the lock f is 
lifted out of the slide /. The man at A may now pull out the 
slide /, which moves the spring contact maker /* and takes 
the lock f 3 out of the quadrant g, to which is attached the 
rod g- leading to the signal. On the handle h being turned 
the pin r on the wheel / travels in the slot g 3 and turns the 
quadrant. Half a complete revolution of the handle h is 
sufficient to lower the signal and the remainder of the revo- 
lution raises it. The wheel / can only travel in one direction, 
a reverse movement being prevented by the pawl j 3 . Fixed 
to the \vheel ; at ; 4 , but on the other side, is a stud which 
raises the arm /* just as the movement is again normal and 
re-inserts lock /*. 

A manual generator is employed to supply power for 
working the instruments and lock. The advantages of this 
method are that the man at B cannot tie down his lever a and 
go to sleep, leaving the line free, and the cost of batteries is 
saved. If ' Track-Circuits " be subsequently added the 
manual generator would supply the energy instead of using 

There are now, or shortly will be, 4,000 miles of single 
line in America, chiefly on the Illinois Central, the Chicago, 
Burlington and Quincy and the Great Northern RR., pro- 
tected by this system, and all this has been installed since 
1904. That the company have reason to be satisfied with 
what has been done is proved by the fact that on 
one section of the I.C.R., 100 miles long, the number 
of butting and rear collisions has been reduced by 
90 per cent., the engine and train crew overtime reduced by 
73 per cent, and the carrving capacity of the section increased 
by from 20 to 30 per cent. 

It would, perhaps, be hypercritical to examine these 
efforts too closely, especially when it is remembered how tar 
they are above what has been done by other American rail- 
road companies. But the Author feels compelled to point out 
their weaknesses, but at the same time to commend the 
action of these companies to grapple with the great problem 
in America of safely operating single tracks. And in con- 
nection with this subject he would like to commend one other 
company, who work the whole of their line on the block 
system, and that is the Chesapeake and Ohio RR. Co., who 
may not have gone as far as the others in the adoption of 
automatic signals and electrical train-staffs, but by the adop- 
tion of the Block System they have made an effort to mini- 
mise that terror of American railroads, viz., the " head-on " 

Dispensing u'ilh Flagmen on Single Lines. 
The Great Western R. have an instrument which has 
been found to give every satisfaction, and by the use of 
which great savings are effected by doing away with the 
necessity of providing flagmen when any engineering work, 
calling for such protection, is being carried out on single 

Rules 248, 250 and 251 of the standard book of rules and 
regulations provide that before any trolley is placed on a 
single line, any rail taken out, or any work done that will 
cause an obstruction, a flagman, with detonators, must be 
sent out in both directions. This is a great demand on the 
resources of a gang, and it means some expense, whilst con- 
siderable time is spent in making arrangements with the 
adjacent signalmen. 

The provision of flagmen may now be avoided on those 
single lines that are operated by the electrical tablet or elec- 
trical train-staff, and the necessary permission for obstruct- 
ing the line more expeditiously obtained bv the ganger where 
the new " Key-Occupation " system is in use. 

The principle of the system may be briefly described by 
stating that in each tablet or train-staff section there is pro- 
vided in a platelayer's hut an instrument (there may be two 
or more instruments, according to the length of the section) 
which contains a key. Though there may be more than one 
instrument in a section there can only be one key, and the 
possession of this key breaks down the tablet or electrical 
train-staff instruments, so that neither tablet nor staff can 
be withdrawn, and, therefore, no train can be admitted into 
the section. If a train be already in possession the kev can- 
not be obtained. 

When the key is withdrawn by the ganger, it is his 
authority for fouling the running line, and he may remain in 
possession as long as he holds the key. 

Fig- 237 shows a ganger obtaining possession of a key, 
and in fig. 238 he is seen with the key in his hand and the 
trolley on the line. 

The use of the keys is, of course, safeguarded in every 
way by the following regulations : 

3. Three gangers are appointed for this branch, working as 
under : 

One from 138 miles 60 chains to 143 miles 50 chains, and using 
Group No. i. 

One from 143 miles 50 chains to 148 mile post, and using Groups Nos- 
2 and 3. 

One from 148 mile post to 152 miles 56 chains, and using Group 
No. 3. 

4. The keys provided may be used in any key box in the group to 
which they belong, but a key belonging to one group must not in any 
circumstances be carried into another group, as it will not fit the key 
boxes of any other group, neither will it afford the ganger the necessary 
protection except when used in the train staff section to which it 

When a ganger has occasion to occupy a portion of the line belong- 
ing to one staff section and then a portion of the line belonging to 
another staff section he must be careful to replace the key appertaining 
to the one staff section before asking for the key of the other. 

When either of the gangers wishes to obtain his occupation key he 
must go to the telephone hut belonging to that particular group in which 
the occupation key has been left, and call up the signalman by pressing 
the call key of the telephone. 

The signalman on hearing the telephone call will reply and take off 

his receiver, and at once inform the ganger he is ready to speak to him. 

The ganger must then make his requirements perfectly clear by stating 

(a| Hut from which message is being sent, giving mileage^ as 

shown in these instructions. 

(b) Length of time occupation is required for. 

(c) Points between which it is intended to do the work. 

(d) When occupation is required for trolleying : The point to 

which the trolley will be run and removed from the line. 
On receiving this information the signalman must communicate with 
the signalman at the other end of the section and agree whether the 
permission may be granted or not. If it is found that occupation cannot 
be granted the ganger must be so informed. If the signalman can 
grant occupation the ganger must be told the exact times between which 
the occupation is given. The ganger must in every casp repeat the tele- 
phone message to show that he understands it. In deciding these times 
_it must be arranged for the occupation to cease ten minutes before a 



P O W E K K i I L TI' ,1 

/ G \ T .1 7. L INC,. 

Fig. 237. 

Occupation of Single Lines, Gt. Western R. 

Fig. 238. 

train is due to leave the next staff station, either up or down, as the 
case may be. The signalmen at both ends of the section immediately 
they have agreed to the occupation must hold down the bell keys of their 
respective electric train staff instruments. 

In cases where two gangers' lengths come in one electric train staff 
section only one ganger at a lime can obtain an occupation key belong- 
ing to that staff section. 

5. Immediately the ganger is informed that he can have occupation 
he must turn the occupation key so that the index shows No. 2, in which 
position the key can be released by the signalmen. 

After turning the key to the No. 2 position the ganger must wail 
until the word " free " appears on a small indicator to ihe right of the 
key, when he must turn the key so that the index shows No. 3. He can 
then take the key ont of the key box and inform the signalman he has 
done so. 

The ganger must keep the occupation key in his possession during 
the time of the occupation. 

6. A ganger, when he has obtained occupation of the line, must so 
arrange his work as to be able to put back the occupation key at the 
appointed time. He may put the key into any key box belonging to the 
group, as the key will fit all the boxes of the group and will restore the 
staff working equally well in either key-box. When putting the key 
back he must turn it to the right until the index shows No. i. Having 
thus restored the occupation key the ganger must call up the signalman 
on the telephone and inform him that he has restored (he key and that 
the line is safe for the passage of trains. 

7. At the time appointed for the occupation to cease the signalman 
who gave the permission to the ganger to occupy the line must listen 
for the telephone call, and on hearing il must reply and receive the 
ganger's message. If all is right for resuming ordinary working he must 
tell the signalman at the other end of the section. The ganger will 
then replace ihe receiver and shut and lock the hut. 

8. In the event of its being necessary to work any portion of the 
line by pilotman, the pilotman on his first journey over that portion of 
the line distributing the pilot working forms must make certain by 
personal observation that the occupation keys are in the key boxes. In 
order that he may be able lo do this duplicate keys of the huts will be 
kept by the signalmen. 

Each occupation key has printed upon it the names of the places 
between which it is available, and it must not in any circumstances be 
carried beyond these places. 

From the foregoing it will be appreciated what economics 
in time and labour may be effected. But there is another 
use for the system. 

There are cases where a foreign railway, a colliery owner 
or manufacturer, or the contractor for a new line or widen- 
ing, has a connection with a single line which has to be used 
occasionally for depositing or receiving wagons of material. 
Ordinarily this would call for a ground frame, if not a signal- 
box, and an attendant signalman, also the conveyance by 
hand of the tablet or train staff from one of the adjacent 
boxes and back again. What is now done is to control the 
lever of the siding points by a key (similar to those already 

noticed) kept in an " occupation instrument." To get on to 
the main line the key has to be obtained from the instrument 
and this breaks down the tablet or train-staff circuit and 
unlocks the siding points. The key is locked in immediately 
the point lever is moved the least, and, therefore, not until 
the points are restored and the key put back can any train 
enter into the section. 

The method of working will be understood from the fol- 
lowing regulations issued to meet such a case as that just 
named : 

1. The contractors will have a daily occupation of the single line 
when practicable between A and B for the purpose of unloading mate- 
rials, &c. 

2. In order to avoid the necessity of sending the train staff for the 
section to the contractors' siding, situated between these stations, when 
the contractors require to bring their engine and trucks out from the 
siding on to the running line, a special key box and telephone have been 
fixed in a hut at the siding. The key of this special box unlocks the 
points of the contractors' siding. 

3. The station master at A must arrange for a competent man from 
the traffic department to be sent to the siding each day. 

4. When the contractors' train is ready to come out from the siding 
the traffic department man must go to the telephone hut and call up the 
signalman at A by ringing three beats on the call key of the telephone. 
The signalman on hearing the telephone call will reply, and take off his 
receiver. He must then be informed what is about to be done, and where 
the contractors will unload their materials. 

5. On receiving this information the signalman must communicate 
with the signalman at the other end of the section, and agree whether 
the permission to occupy the line may be granted or not. If it is 
decided that the occupation cannot be granted ihe man at the siding must 
be so informed. If the signalmen agree lo granl ihe occupation the man 
at the siding must be told the exact times belween which the occupation 
can be given. The man at the siding must repeat the lelephone message 
lo show lhat he understands it. In deciding these times it must be 
arranged for the occupation to cease ten minutes before a Irain is due 
lo leave the next slaff station, either up or down, as the case may be. 
The signalmen at both ends of the section, immediately they have agreed 
to ihe occupation, must hold down the bell keys of their respective electric 
train staff inslrumenls. 

(i. When the key has been taken out the needles of the current in- 
dicators in the staff instruments will assume the vertical position, which 
will indicate to the signalmen lhat the key has been removed, and they 
can release their bell keys. 

7. Immediately the man in charge cf ihe siding is informed (hat it 
is right for the contractors' train lo come out from the siding he must 
turn the key in Ihe instrument so lhat the pointer points lo No. 2 
position, in which position the key can be released by the signalmen. 
After turning the key to the No. 2 position the man must wait until 
the word " free " appears on a small indicator to the right of ihe key, 
when he must turn the key so that the pointer poinls to Ihe No. 3 
posilion. He can then take the key out of the key box. 

X. The siding points can then be unlocked, and the traffic department 
man must accompany ihc j Irain and keep the occupation key in his 

SINGLE Ll\ES, KEV O C C f 1> A T 1 O X . 


possession until the train has again been placed clear in the sidmg, and 
no obstruction has been left on the single line. 

9. When the man in charge of the special key has satisfied himself 
that no vehicle or other obstruction has been left on the running line, 
and when he has properly locked up the siding again, he must go back 
to the hut and put the key back in ihe inslrument. When doing this he 
must turn the key to the right until the pointer points to the No. I 
position, and having thus restored the key he must call up the signalman 
en the telephone and inform him that he has restored the key, and that 
the line is safe for the passage of Irains. At the time appointed for the 
occupation to cease the signalman at A must listen fcr the telephone 
call, and on hearing it must reply and receive the message. If all is 
right for resuming ordinary working he must tell the signalman at the 
other end of the section. The man at the hut will then replace the 
receiver and shut and lock the hut. 

10. In the event of its being necessary to work this section of line 
by pilotman, the pilot man on his first journey through the section dis- 
tributing the pilot working forms must make certain by personal obser- 
vation that the occupation key is in the box. In order that he may do 
this duplicate keys of the hut will be kept at A and B. 

Yet another use, and this is one that will appeal to some 
British Rs., and still more so to foreign railways, is that 

where there are intermediate sidings or passing places trains 
that have work to do there can be provided with a key at 
the signal-box at the entrance to the section which will gain 
access for the train to the siding or passing place. When 
the train is " inside " the trainmen will insert the key in an 
" occupation instrument " near the points (which would then 
be locked up), and this would re-open the tablet or train- 
staff section and allow other trains to pass. When the time 
arrived the key could be again withdrawn and the train pro- 
ceed en its journey, when the key would be placed in a corres- 
ponding instrument at the other end, or the train might 
return to the entrance of the section, and the restoration of 
the key would re-open the line. 

Fi - 2 39 ' a view of the interior of the " occupation 
instrument," and figs. 2^0 to 242 are diagrams of the con- 
nections between the tablet and train-staff instruments and 
the "occupation instrument," and which will be clear from 
the following description : 

When the occupation key is in possession of the per- 
manent-way ganger or other person it is not possible to 
obtain a staff or tablet. To accomplish this one or more 
instruments each having a slot into which a key may be in- 
serted are provided intermediately between two electric 
tablet or train staff instruments in a section. The key, on 
which is engraved the mileage of the section over which it 
affords protection, is normally locked in one of these occu- 
pation boxes. The apparatus is so constructed that the kev 
may be released by the signalmen at the ends of the section 
only when all the staffs or tablets are in their instruments. 

The occupation box itself and the method of working it 
vary somewhat, depending on whether it is required to 
work in connection with the staff or tablet system and also 
what function it is required to perform. In the most general 
method of working a number of boxes are provided with 
only one key ; this may be inserted in any box of the set. 
The kev can be withdrawn from the box in which it is locked 

Fig. 240. Connections for Instruments; Occupation of Single Lines, <i.\Y.R. 


and replaced in the same or in any other of the remaining 
boxes of the group. The method of working is shown in fig. 
240. The occupation instrument consists of an electric lock 
in combination with an automatic switch, the latter, which is 
of the commutator type, being actuated by the movement of 
the occupation key. The electric lock is of a compound type 
having two independent coil windings. Unless there be a 
current in each of these and in the proper direction the pole 
pieces do not energise and consequently the lock or catch 
rendering the key relcasable is not operated. Ordinarily 
these coils do not form any part of the electric train staff 
circuit but arc switched in automatically when required. 

The electric train staff apparatus and occupation boxes 
are worked upon two line wires. It will be seen that the L 
terminal of the left hand electric train staff instrument is 
connected with the L terminal of the right hand staff instru- 
ment through the springs A and A 1 of the occupation box 
switch (two boxes arc shown, but any number, say a dozen 
or more, would be quite similarly connected). The return 
line from the E terminal of the right-hand instrument is con- 
nected through the occupation box switch springs B and B 1 
to the E terminal of the left hand instrument. So far as the 
train staff instruments arc concerned this circuit does not 
differ from that ordinarily in use except that a return line is 
used instead of earth, which is sometimes possible. 

The occupation key is capable of taking up three different 
angular positions when in an occupation box, viz. : 

1. (Normal.) Key locked in box. The electric train staff 
circuit is intact and staffs can be withdrawn as usual. 

2. The key is always free to be turned to this position 
from No. i, but beyond this it cannot be further turned 

Special CommuMorSniich 

Electro- Magnet Lock Controlling 
Special Commutator Strrtch 


unless an electric lock which engages with a stop on the 
disc carrying the commutator is released. In this position 
the automatic switch divides the electric train staff connect- 
ing the line and return wire from one end of the staff section 
to one coil winding of the electric lock, and the wires from 
the other end of the section to the other coil winding. If 
now the tapper keys of the staff instruments be depressed, 
current will be sent from each end of the staff section into 
the electro-magnet coils. The electro-magnet being con- 
structed on the closed circuit principle, poles are not created 
unless both currents are present and in proper phase, that 
is to say, all the staffs must be in the staff instruments. 
The lock is now released and the key can be turned to 
position 3. 

3. The key can now be withdrawn from the box. In 
this position the automatic switch inserts a cross in the 
lines, that is, the L terminal of one staff instrument, instead 
of being connected direct to the L terminal of the corres- 
ponding instrument at the other end of the section, is con- 
nected to the E terminal. In this way the staff instruments 
are thrown out of phase during the time the occupation key 
is not locked up in one of the boxes. The insertion of the 
cross at the box from which the key is withdrawn is in fact 
equivalent to a staff being out of the staff machines so far 
as the latter are concerned. When the occupation for which 
the key was obtained is completed, the key can be replaced, 
either in the box from which it was removed or in one of the 
other boxes of the section, when the staff working is again 
restored to normal. The object of this arrangement of only 
one key to a number of boxes is to enable the ganger to 
obtain the key at one point so as to enable him to trolley 
to another. It is necessary to afford the ganger means of 
communication with the signalman at the staff stations in 
order to arrange the occupation. This is quite simply ac- 
complished by means of a superimposed type of telephone 
phonopore or condenser pattern connected across the two 
staff wires. The telephones work independent of the staff 
and do not affect it in any way, but switches arc usually 
provided, so that the telephone is only in operation when 
actually required. 

Fig. 241 shows an alternative method of working the 
boxes ; here the occupation instruments have no actual con- 
nection with the staff line, but are controlled on a special 
line of their own. At one of the staff stations a special 
commutator switch and electro-magnet locking the same is 
provided. In position i of the commutator the staff circuit 

Electric Train Staff Line 

Special tine 

i i 

Fig. 241. Special Wire for Connections 



connections are normal. To release the key the special com- 
mutator switch, worked with a handle, must be turned to 
position 4. The commutator is first turned to position 2, 
where the electro-magnet lock prevents further turning 
movement. If current is now sent from the other end of 
the staff section and all staffs are in the instruments, the 
polarised relay will be operated to close local circuit of 
electro-magnet locking special commutator switch. 

In Xo. 2 position the polarised relay is automatically 
switched in. Electric lock is now released and special com- 
mutator can be turned to position 4, in which it connects a 
battery to the special occupation line. The occupation in- 
struments in this method are rather more simple, consisting 
of switch with three springs and an ordinarv horse shoe 
type electro-magnet lock. If the key in the occupation box 
is now turned to position 2 current is received from special 
line through its electro-magnet coils, which release lock, 
allowing key to be turned to position 3, in which position 
it can be withdrawn from the box. The staff instruments 
are now totally inoperative, as when the special commutator 
is turned beyond Xo. 2 position the staff line, which is 
obtained through contacts on the special switch, is discon- 
nected. To restore the system the occupation key is re- 
placed and turned to Xo. i, here it connects a battery with 
the special line so as to send a current to the staff station 
(for this purpose the telephone battery gives all the power 
that is needed). The special commu- 
tator is turned to position 3, where it 
is again locked, but if the key be in 
one of the boxes, the polarised relay 
will operate bv virtue of current 
recehed from occupation box battery, 
local circuit of special commutator 
switch will be closed, lock released, 
and switch can be turned to position 
i , where everything is again right for 
normal working. 

| Normal Connections 
of Commutator 
N? I Position 

The disadvantage of the latter method of working is that 
it introduced additional apparatus at the staff station, and 
also relies on a batterv for resetting the apparatus, but it has 
the advantage that the staff line is left intact, all the switch- 
ing being done at the staff stations. It is also possible by 
this arrangement to release two or more keys appertaining 
to different groups of instruments, so that two occupations 
may be given in the electric train staff section at one time, 
which is not possible directly by the first described arrange- 

For applying the o-.-cupation system to Tver's electric 
train tablet apparatus, the principle of working just described 
is adopted. As in the case of the train staff a special electro- 
magnetically-controlled commutator is necessary for operat- 
ing the occupation instruments, which are exactly similar 
to those described in the second method of working them 
with the staff. The modus operandi is as follows : The 
commutator is turned to Xo. 2 position and a current sent 
from the other end of the section, which actuates polarised 
relay in tablet instrument. This, in the ordinary course of 
things, would close the circuit of the tablet locking-magnet 
and release tablet, but is arranged by commutator to close 
local circuit of electro-magnet controlling commutator in- 
stead. Lock being released, commutator is turned to posi- 
tion 4, where it releases occupation key in the manner pre- 
viously described. For resetting, current is received from 
the occupation box, and when commutator is turned to 
position 3 this current operates a separate relay which 
in turn works control magnet of special commutator, 
allowing latter to be restored to normal position. The 
tablet instruments are rendered inoperative while the 
key is absent from the boxes owing; to the line wire con 
nection being disconnected in the special control switch, 
and as additional safeguard the battery is also 
disconnected from the tablet instrument at the 
controlling end, so that even if the line wire 

in N?l. fysit' 
'Key in ' 

/V? J. Position 

'Key out' 

'Key out ' 

N?3. Position 
'Key out' 

TaHet Line 

- ' 

^_ Speca'.Baltery 

Corrections of 'Commutator 
inffrrrvtjiafe position 


Fig. 242. Connections for Tablet Instrument. 



were by an unforeseen chance to remain intact, current could 
not be sent out, neither would current be available for re- 
leasing the locking magnet which frees the tablet. 

If it is necessary to provide only one box in an electric 
train staff section from which a key can be taken out and 
replaced for the purpose, say, of working a contractor's 
crossing on the ordinary line level, all that is required can 
be done on the one wire of the train staff circuit with earth 
returns. In this case the staff line is joined through the 
contacts of the occupation box normally. When the key 
is turned to position 2 one coil winding is connected to one 
side of the electric train staff line wire and earth, and the 

other coil winding with the opposite side of the electric train 
staff line wire and earth, current being sent into the coils of 
the instrument by the depression of the tapper keys at the 
end of the electric train staff section. Lock is released and 
key turned to position 3 and withdrawn and utilised for 
releasing points, etc. In position 3 the electric train staff 
circuit is totally disconnected, so that no staff can be 
obtained. This holds good until the key is replaced to No. 
i position and line connection made intact. 

Tver and Co. are the sole licensees for the manufacture 
of the instruments, etc. 

Fig. 509. Front view of " AH-Klcctric " Locking Frame at Didcot, O.U'.R. (sec page 277). 



AUTOMATIC signalling is, and always has been, regarded 
as a substitute for the block system and offered as an alter- 
native to the equipment* of lines with signal boxes and 
mechanically operated signals, with the attendant staff of 

Though it is not purely an American idea, it was first 
put to practical use and subsequently developed in America. 
When first used the main idea was the protection of trains 
from following ones and no attempt was made to provide 
for interlocking or the protection of shunting operations. 

Automatic signalling has found congenial ground in 
America. The lines there were built to create traffic, and 
it was not until after they had been built and in use for 
some time that the need arose for their protection. Such 
protection as is given here or on the Continent was out of 
the question in America. The stations were many miles 
apart and one of the troubles was to get more trains over 
the line and therefore to provide some signalling in between 
the stations. But the fixing of a signal box, signals, etc., 
and the stationing of a man or men was not to be thought 
of, not onlv on account of the cost, but because no men would 
accept the position of a signalman at a post outside the 
bounds of civilisation. At stations, too, the fixing of signal 
boxes and signals was again out of the question. The 
points were worked from the ground and there was con- 
sidered to be no necessity for concentration and interlocking. 
Men, too, were scarce, and labour dear. 

A method, therefore, whereby trains signalled themselves 
appealed to American railroad officers. The progress of 
automatic signalling was slow, but so was any other form 
of protection, and it was not until the early eighties that 
any advance was made. The officials on American lines 
began then to see, as their British brethren had done 10 
or 15 years earlier, that signalling spelt safety and speed. 

The type of signal and its method of operation has 
grown. Originally it was a disc that was exposed when in 
the "on" position, and was withdrawn when at "clear." 
These were operated by magnets. Another original form 
was a disc operated by clockwork. These eventually gave 
place to a semaphore signal, where the arm was lowered 
by a signal rod operated by an electric motor. But prior 
to this there had been automatic signals of the semaphore 
type that were worked by compressed air under the \Yest- 
inghouse system. The air was conveyed by pipes from a 

reservoir \\hich, in many cases, was 20 or 30 miles away. 
About 1902 the Hall Signal Co. brought out their electro- 
gas signal, which was an interesting departure. 

The means of operation have also changed. Primary 
batteries were originally used for actuating the discs and 
working the motors, except in those cases just referred to, 
where the signals were lowered by air under pressure. The 
electro-gas signal has carbonic acid gas stored under pres- 
sure in cylinders as its motive force, the valves of the gas 
cylinder being opened and closed by electrical relays, in 
the same way as the compressed air is controlled. By the 
use of the gas cylinders there is no need to run a pipe 
line. A change is now being made as regards batteries. 
Secondary batteries are employed and a line wire run from 
the power station. Portable secondary batteries are also 
used which are charged at a power house and taken out to 

the signal. 

The original automatic signals indicated " clear " nor- 
mally and only went to danger when a train passed. After 
the train had gone the required distance the signal was 
released and went to " clear " automatically. This was 
considered inconsistent, seeing that the standard position of 
signal arms is "on," and consequently the "normal 
danger " system was introduced, whereby automatic signals 
stood normally in the " on " position. After being lowered 
for a train to pass them they were put to danger in the 
usual way and remained there, even after the train had gone 
out of the section, or sections, ahead, and were lowered 
by the approach of a train, providing that the line was clear. 

Originally the signals were raised and lowered by cur- 
rents set up by electrical treadles or other forms of contacts, 
and no " Track-Circuits " were used. The advent and devel- 
opment of " Track-Circuits " has changed all that, and such 
are the complement of all automatic signal installations now. 

Distant signals were not, or only rarely, at first used in 
America. To provide for trains overrunning a home signal 
it was customary to give what is known as an "overlap," i.e., 
a space of certain length past the next stop signal which had 
to be clear before the stop signal in the rear could be low- 
ered, so that the lowering of a stop signal gave a clear 
section to the next stop signal and a bit more. 

" Wireless-Circuits " is the name given to an ingenious 
arrangement brought out by the Union Switch and Signal 
Co., by which it is possible to actuate a distant signal in 


01 V K R RAt t. W A F S / G N A L L 1 N G . 

sympathy with its corresponding home signal without run- 
ning- a line wire from the home to the distant signal. The 
necessary contacts are made through the " Track-Circuit " in 
the rear of the home signal. 

Mr. Richardson, of the Cic. tic Signaiix electric/lies pour 
Clieminx tie jcr of Paris, has also designed a " Wirclcss- 
Cireuit " system. 

' Three-position " signals are peculiar to certain rail- 
roads in America. When horizontal it indicates stop as 
usual. If lowered to an angle of 45 it indicates that the 
section immediately in advance of the signal is clear, but 
if lowered through an angle of 90 into the vertical position 
it indicates that the next two sections are clear. By this 
means the provision of distant signals is rendered unneces- 

Part of the complement of automatic signalling installa- 
tions are Indicators to show the approach of trains and 
whether the section in advance is occupied ; Switch Locks, 
which break the " Track-Circuit " current in the event of any 
switches being opened or not properly closed ; Slots, by 
which the " Track-Circuit " controls mechanical signals and 
thereby converts them into semi-automatic; Insulated Joints 
for dividing the rails into sections; and Relays. 

Long-burning Signal Lamps are more or less directly 
connected with the adoption of .automatic signals. Such 
lamps, instead of requiring to be filled, trimmed and cleaned 
every day, only need attention once a week. The economies 
in oil, labour in fetching in and taking out the lamps for 
trimming and recharging, and their longer life are very 

Equipping Electrically Worked Railways with Track- 
Circuits was an altogether different problem, which arose 
about 1900. to that of providing them on railways worked 
by steam. The difficulty was overcome, and now the Boston 
Elevated, the San Francisco North Shore, the Intcrborough 
of New York, the' Long Island, and the \Ycst Jersey and Stja 
Shore arc amongst the railways equipped in America, and 
the Metropolitan District, Great Northern and City, Lanca- 
shire and Yorkshire (near Liverpool), Baker Street and 
Waterloo, G.N., Piccadilly and Brompton and the Charing 
Cross, Euston and Hampstead railways in England. 

The frequent, quick and punctual services on the New 
York Interborough R. is dependent upon automatic signal- 
ling, and it is an eloquent fact that on that line there were 
during the month of February, 1905, signal and automatic 
stop movements to the number of 4,206,720, and only 13 
failures, or an average of one failure for 323,594 movements. 
The average number of movements ' has increased to a 
monthly average of 5,282,028, but the rarity of failures is 
still most remarkable. 

The speed, regularity and safety obtained from automatic 
signals on steam railroads is also astonishing. 

In America, where there is so much automatic signalling, 
there are apparently very few failures, and especially of those 
on the danger side. It is to be regretted, though, that 
some companies have to adopt surprise checking in order 
to see that automatic signals are obeyed. 

Altogether, automatic signalling has been brought to a 
fine state of perfection, and particularly is this so in this 
country, where undoubtedly the railway companies who have 
taken up the system, and the signal contractors interested, 
have certainly improved upon American practice, and this 
is saying a great deal. 

Automatic signals arc an expensive item, and where 
mechanically operated signals already exist and in Great 
Britain they are in use on every line open for traffic their 
introduction can only be justified when the expense can be 
recouped by economies effected by signal boxes being closed 
and signalmen dispensed with. These cases are, however, 
rarer than is generally imagined. The average cost in Eng- 
land, including fitting the "Track-Circuits" and running line- 
wires, is about ,"100 per arm or ,"200 for a two-arm signal. 
The maintenance charge is also high, varying, according 
to the report presented by Mr. Platt to the International 
Railway Congress in 1905 at Washington, from 12 to 
over 20 per arm a year. Then each signal has to bear a 
charge of about 2 for lighting, and there are also fogging 
expenses. In the case of new lines, as yet unsignalled, it 
is no doubt much cheaper to provide automatic signals in- 
stead of equipping the road with signal-boxes, signalmen 
and mechanical signals. But new- lines are most uncommon 
in England, and the question must always be considered 
in relation to existing equipments. These have been paid 
for, and naturally there must be some good reasons for 
installing a new system. Where there are signal-boxes at 
which there is little shunting, and where the chief duty of 
the signalman is to work the block instruments and signal 
trains, these men may, if automatic signals be provided, 
be taken out of the signal-boxes and need only go in when 
any shunting has to be performed. Such cases will appeal 
to railway managers, but it takes a good many such 
economies to pay for automatic signals in Great Britain. 

For example, assume Automatic Signals to be installed 
on a British railway over a length of 20 miles of double 
line with signals one mile apart. This would require forty 
2-arm signals or 80 arms in all, which would cost, say, 
;,8,ooo. The annual charges would be approximately 
;, 1,980, made up as follows: 

Interest at 4 per cent, on ,8,000 
Sinking fund for renewal in 20 years 
Lighting 80 signals at 2 
Fogging 40 distant signals at _."i 
Maintenance of 80 signals at ,"15 

... 320 
... 260 
... 160 
. .. 1,200 

The average pay of a signalman is 22s. per week ^.'57 
a year to which may be added .'3 for his uniform, holiday 
pay, etc., or ^,'bo in all, so that it would require 33 signal- 
men to be dispensed with to justify this expense. 

This estimate and conclusion applies to what may be 
termed a wholesale adoption of automatic signalling, and 
even such cases may be modified under certain conditions, 
as, for instance, on the L. and South Western R. between 
VVoking and Basingstokc, where the automatic signals and 



the "Track-Circuit " are part of the equipment of a power 
signal and interlocking plant. There, the power for work- 
ing the automatic signals is generated by the same means 
as the power for operating the points and signals at stations, 
and this considerably reduces the cost of maintenance, which 
is the leading item in the expenditure. 

Xor does the comparison between automatic and 
mechanical signals apply to Tube railways or such railways 
as the Metropolitan and the Met. -District. Firstly, because, 
being electrically operated, power is cheap, and, secondly, 
because it is physically impossible, and financially, too, to 
provide signal-boxes and signals at such short intervals 
apart as the headway of traffic calls for, and which condi- 
tions automatic signals so readily and completely meet. 

The Author has been much surprised at the low working 
and maintenance costs of the automatic signals and power 
plants on the Tube railways of the I'nderground Electric 
Railways Co. of London. 

The total mileage of single line in the three tube railways 
is 4 1 '68 miles and the average number of signals is I5"85 per 
mile. During a week chosen at random there were a total of 
1.538,282 "round movements," that is, the lowering and 
raising of a signal and the reverse and normal movements of 
points. During a period of 13 weeks there was an average 
of 14 failures per week of either points, signals, train stops, 
locking frames or their connections, and which were all 
failures on the side of safety. This gives an average of i 
failure in 109.877 movements. The loss of time to trains 
debited to these failures works out at 15 seconds per 1,000 
train miles. These figures speak for themselves. The cost of 
operation is equally satisfactory. The total cost per pas- 
senger train-mile works out at 0^457 of a penny, and the cost 
per mile of single track is ."4 35. 6'68</. Included in these 
figures are the cost of the power (as charged by the Chelsea 
Power Station); material used in repairing, cleaning, etc., 
the signals, points, facing-point locks, locking bars, train 
stops, and locking frames ; the lighting and cleaning of the 
signal lamps and the wages of the signalmen, repairmen, 
linesmen and inspectors, and all this, divided by the number 
of signals, works out at 5.*. y22ti. per signal. 

In Great Britain the greatest use of automatic signalling 
on ordinary steam railways will be found in dividing long 
block-sections and as an alternative to, and a substitute for, 
intermediate block posts. 

It is only fair to put to the credit of automatic signals 
that they have further advantages than those of saving 
signal-boxes and signalmen. They effectively guard against 
errors in block working, due to trains, or parts of trains, or 
vehicles left in a section or standing foul. Further, thev 
can be placed as close together as the traffic demands, and 
so add to the carrying capacity of the line. 

There is, however, one objection to placing signals close 
together, which is the strain placed on a driver by having 
continually to look out for signals. 

On most railroads in America an automatic signal arm 
is distinguished from a mechanically worked arm by the end 
being pointed. Nothing of the sort has been attempted in 
Great Britain. This remark creates the opportunity of re- 

ferring to a difficulty drivers may meet with under certain 
conditions. It may be assumed that a certain signal is semi- 
automatic, and acts as, say, the starting signal for a signal- 
box during the day (when the box is open) and is controlled 
by the " Track-Circuit " in advance and is purely automatic 
at night, when the signal-box is closed, the signalman before 
leaving pulling off his slot so that the signal responds to the 
' Track-Circuit " entirely. 

Now, it is a rule that mechanically operated signals, or, 
to be correct, starting signals, must not be passed at danger. 
On the other hand, automatic stop signals may be passed 
at danger after the driver has stood for a certain length of 
time. In the event of such a signal (a combined mechanical 
starting and automatic stop) being at danger, it would be 
difficult for the driver to know whether it was being worked 
by the signalman, in which case he must not pass it at 
danger, or whether it was being worked from the "Track- 
Circuit," in which case he may pass it after waiting the 
required interval. It is easy to imagine conditions under 
which this question might arise when a decision could not 
be arrived at because the signalman could not be seen in 
the box. 

' Xonnal Danger" and " Xornia} Clear." 

The original signals in America stood normally at 
" clear " unless there was a train in the section. In 1891 
the circuits and details were re-arranged by the Hall Signal 
Co., so that the signals could remain at danger, after being 
placed in that position by a train entering a section, until 
they required lowering for a second train. This alteration 
was regarded with favour by many officers, as automatic 
signals standing normally at danger agreed with the normal 
position of mechanical signals. 

This favourable opinion, however, was not shared by all 
railroad officers, and it is still a thorny point in America. 
Roth systems without doubt have their good points, but 
more is made of the matter than there is any need for. In 
the Author's opinion little would have been heard of the 
subject had it not been for the fact that the "normal- 
danger " arrangement was the subject of letters patent 
owned by one signal company, to whom all who favoured 
" normal-danger " had to apply. 

In Great Britain the patent has expired, and railway 
companies can please themselves, and when the patent has 
run out in America the Author anticipates that the differ- 
ences of opinion will be modified. 

All the automatic signals in Great Britain except those 
on the North Eastern R. are " normal-clear. " 

In America, opinion is fairly evenly divided as to the 
merits of each system. The advocates of " normal-clear " 
claim that the electrical connections are fewer with this 
svstem than with the " normal-danger," and that it is easier 
to see whether the signals are in working order. This latter 
is a point of some importance, when it is remembered what 
lengths of line they have in America between stations and the 
large mileage a linesman has to look after. \Yith " normal- 
clear " signals a linesman is able to stand at the rear of a 
car and see if the signals go to danger behind his train, and 
on a long straight stretch, if the second signal behind him 



clears when his train has gone into the second section. He 
can also watch the behaviour of the signals in the opposite 
direction, which should be standing "off," if no train is 
in the section, and when a train passes him he is able to see 
if the signal immediately in the rear has gone to danger 
and those behind have again cleared. 

On the other hand, " normal-danger " signals corres- 
pond with standard practice, and in case of failure due to 
frost or snow they are more liable to have the arms frozen 
in the danger position, as that is the position they are in 
for the greater part of the time. "Normal-clear" signals 
for the same reason are liable to be frozen or stick at " clear." 

A further advantage is that no battery power flows when 
signals are at danger, as the connections are weighted so 
that the arms assume that position in case of failure. Con- 
seauently power is reauired to hold the arms at "clear," 
and the battery is therefore always flowing when signals are 
"off." With "normal-clear" signals the amount of bat- 
tery thus consumed is small, but it is eliminated where sig- 
nals are normally at danger. Another advantage of the 
"normal-danger" is that platelayers working on the line 
can be warned by the lowering of the arm that a train is 
approaching, but no such security is afforded by a signal 
that stands normally "off." 

Consequently the two systems of working, that of 
" normal-clear " on the L. and South Western R. and the 
" normal-danger " on the North Eastern R., will be followed 
with interest. 

As to whether the Rules and Regulations Committee of 
the Railway Clearing House, who are watching the question 
of power and automatic signalling, or the Board of Trade, 
will be able to decide whether automatic signals shall be 
normally " on " or " off " it is hard to say at this moment, 
but it is to be hoped that a decision may be announced so 
that railways and contractors may not have the objections 
ever with them that come from two systems of working. 

Personally, the Author is in favour of " normal-clear," 
and this, on the grounds that an automatic signal is different 
to a mechanical signal, which has the brains of the operator 
behind it, and acts instead of his verbal orders to stop or 
proceed. A mechanical signal, when lowered, tells the 
driver not only to proceed, but that his approach has been 
accepted by the man at the next signal-box. An automatic 
signal, when lowered, simply says " thc-section-is-clear." 
If, then, an automatic signal is only an indicator as to the 
section ahead being clear, it surely is only consistent that 
it should be at " clear " when the section is clear, and thus 
be "normal-clear." 

" Overlaps." 

On the subject of " overlap " there is not the same dif- 
ference of opinion. 

In America, as has already been said, it was not the 
custom originally to provide distant signals, and so an over- 
lap had to be provided to allow for an over-run in case a 
driver passed a stop signal at danger. Such overlaps have 
been continued by most companies, even although distant 
signals have been provided. 

In the United Kingdom an " over-lap " appeals to some 

railway officers, so as to be in line with our Block Regula- 
tions, Clause 4 of which provides that " the line must not 
be considered clear, nor must a train be allowed to approach 
from the signal-box in the rear .... until the preceding 
train has passed at least a quarter of a mile beyond the home 
signal. " 

The object of this regulation is to guarantee that a space 
is provided between two trains in case the second over-runs 
the home signal, and in automatic signalling the over-lap is 
put in for the same purpose. It cannot, however, be re- 
garded complacently, as it is really a confession of weak- 
ness, and an admission that drivers run by stop signals at 
danger. On the other hand, it must be granted that it is not 
cheerful to think that only the thickness of a signal post may- 
separate the tail of one train from the head of another. For 
after all this is what it resolves itself into without an 
" over-lap " a second train can leave A when the tail of the 
first train is immediately past the stop-signal at the en- 
trance to the B C section. With an overlap the second 
train must have travelled some distance into the B C sec- 
tion before the signal at A can be lowered. 

On the North Eastern R. an over-lap of 400 yards has 
been provided, so that a second train cannot enter a section 
unless that section and the first 400 yards of the next sec- 
tion are clear. 

The Metropolitan District and the Underground Electric 
Co. 's Tube railways have an overlap of 4ooft. This short 
length is sufficient, although there are no distant signals, 
as the signals and trains arc equipped with an automatic- 
stop which applies the continuous brake, which will pull up 
any train in that distance. 

On the other hand, on the L. and -South Western R. no 
"over-lap " is provided, as it is considered that the distant 
signal in the " on " position should be sufficient warning 
and give ample time to a driver to pull his train up before 
reaching the stop signal. This, after all, is consistency, 
and the L. and South Western R. officials are to be con- 
gratulated on adhering to their opinions. The Board of 
Trade have passed the arrangement. Briefly, then, the in- 
stallations on the North Eastern R. and the L. and South 
Western R. may be regarded as object lessons in these two 
respects the former has " normal-danger " with an "over- 
lap," and the latter " normal-clear " without an " overlap." 

The " over-laps " on the Interborough R. of New York 
have been arranged on a scientific basis to suit the gradients 
and curves, and to provide for the maximum speed and 
braking conditions. Every portion of the line has been 
dealt with separately, according to the gradient and curve. 
A speed of 35 miles per hour was assumed, and 50 per cent, 
was added to the braking distance as a margin of safety. 
The result determined the distance from one signal to the 
next, but before a signal can be lowered the line must be 
clear up to the second signal in front of it. Therefore each 
signal has an overlap in front of it extending up to the next 
signal, and each block section contains two stop signals. 
In other words, before A can be lowered the line must be 
clear to C, but if B be at danger the man has the distance 


from B to C in which to stop, and this length must be un- 
occupied or signal A could not have been lowered. This is 
one feature of interest as regards " over-laps," and the other 
is that there is no hard and fast rule as to the length of the 
" over-laps," as the gradient and curve determine that. 
Automatic. Train Slops. 

On some electrically operated railways, e.g., the Boston 
Elevated, the New York Interborough, the Philadelphia Sub- 
way, the Metropolitan District, the Baker Street and \Yater- 
loo, the Charing Cross, Euston and Hampstead and the Picca- 
dil'y end Brompton, the signals are equipped with an automatic 
stop which opens the continuous brake on the train in the 
event of a train passing a stop signal at danger. This is 
an extremely useful safety appliance, and particularly ap- 
plicable to underground railways, where there is no proper 
view of the line and where the sections ;;rc short. 
Automatic Signalling in Great Britain. 

Owing to the universal adoption of automatic signalling 
in America it is generally assumed that the idea of a train 
automatically signalling itself originated in that country. 
There is, however, a doubt about this. It has been stated 
that the first application was made at Hartford, Conn., 
U.S.A., by Thos. S. Hall in 1866, but in October, 1860, 
a British patent was filed by \Yilliam Bull by which " Track- 
Circuits " and Cab-Signals were anticipated. It was therein 
proposed to make the rails conductors of the current, so that 
the train could at certain points be communicated with, or 
the guard or driver could communicate with another train 
or station. Portions only of the rails would be insulated, 
and an indicator on the engine would show when those por- 
tions were being traversed and also register the distance 

Then, in the early days of railways, and particularlv 
from 30 to 40 years ago, there were several schemes pro- 
posed, and some adopted, for clock-work signals. So long 
ago as 1850 Mr. Tver developed a scheme whereby the pas- 
sage of a train out of a section unlocked the signals at the 
entrance to that section, and in 1872 Mr. Sykes submitted 
to the Metropolitan District R. a proposal for automatically 
signalling their line by a series of electrical rail contacts 
which caused a disc signal to be put to " danger " as a train 
passed it and the disc signal immediately in the rear to be 
put to " clear." The signals consisted of a fixed lamp with 
a movable screen with red glass, so that a red light was 
shown when the signal was at danger and a white light when 
the signal was cleared. Such an arrangement has been 
adopted on the Metropolitan R. of Paris, and is described in 
a succeeding chapter. Nothing came of these proposals, and 
no automatic signals, except experimental ones, subsequently 
removed, were adopted in Great Britain until 1893, and a 
few years later on the Continent. Various causes contri- 
buted to this so far as Great Britain is concerned, and these 
apply generally to the Continent also. One leading cause 
was that, wherever a signal-box was provided, there were 
generally some points to be operated. Even at outlying 
places where there was neither a station nor a side-track 
it was customary to provide a cross-over between the main 

lines. Signalmen were required to advise the line in ad- 
vance of the approach of trains, to intimate the class of train 
(passenger or goods, express or local, special, etc.), to take 
any action necessary should a passenger be giving signals 
of alarm, any doors be open, anv vehicle on fire, any axle- 
boxes be smoking, any load shifted or any train broken loose. 
They were useful in sending intimation of an accident and 
arranging for all traffic to be worked on one line. They 
were also required to advise trainmen how following trains 
were running, and to give instructions for one train to shunt 
to allow a more important train to pass. It must, of course, 
be remembered that the Train-Dispatcher does not exist in 
England ; the signalmen know how trains are running and 
act on their own initiative. 

These points were well brought out by Mons. Margot in 
his paper on Automatic Signals read at the International 
Railway Congress at Washington in 1905. As evidence of 
the services that signalmen render by reporting anything 
wrong with a passing train M. Margot quotes the following 
statistics : 

On the Paris, Lyons and Mediterranean R., between I-aroche an ! 
Dijon, there are 3(1 signal-boxes and an average of 80 trains daily in 
both directions. During the year iqoi these men stopped 07 trains for 
the following reasons : Lights of tail-lamps out, 4 ; defective couplings, 
28 ; jammed brakes and hot axle-boxes, 2ti ; carriage doors open and 
defective loading, 20 ; miscellaneous, iq. 

On a section of the French State R., 20 miles in length, where the 
daily average number of trains is 76, there were 70 trains stopped 
during the year IQOI, oo of which were because the tail-lights were out. 

On the Orleans R., with 14 signal-boxes and no trains per day, 
there were 70 trains stopped during the same period. 

For such contingencies and purposes as those related the 
human agency only, and not a machine, is available. Again 
there is not the same need as in America. From the begin- 
ning of railways there had been men whose sole duty it was 
to protect the movements of trains. These were originally 
called " policemen," who displayed red or white flags. \Yhen 
semaphore signals were introduced they operated them, and 
subsequently they worked the Block-system, so that the ques- 
tion of providing automatic signals as an alternative to pro- 
viding signal-boxes and signalmen never seriously arose, as 
the men were already there. Labour, too, was, and is, cheaper 
in England, and is cheaper still on the Continent, and as sta- 
tions and sidings are close together (as compared with the 
vast stretches in America), so there was no difficulty with 
regard to signalmen, fitters, and others having to live miles 
from any civilisation, as would often be the case in America 
if the railroads there were protected by signals worked by 

It was also natural that British railway men, belonging as 
they do to such a conservative nation, should be slow to 
move and to abandon old, well-tried methods, which have 
stood the test of time, in favour of those which had not then 
passed out of the experimental stage. The human agent, it 
was true, was expensive, but it was reliable. 

But one of the leading factors that have militated against 
the use of automatic systems has undoubtedly been the con- 
trol exercised by the Board of Trade. Such oversight is un- 
known in America, and consequently there is greater freedom 
to adopt new ideas, besides a natural aptness to secure labour- 
saving appliances. But in Great Britain no new system of 
signalling may be adopted until it has been approved by the 



Board of Trade, and while they are always willing to con- 
sider any proposals, and will give encouragement to any 
practical idea and this is particularly true of the present 
generation of inspecting officers yet they naturally arc 
cautious and will not give the mark of approval to any new 
method of working until it has stood the severest possible 
tests under all the various conditions likely to arise. 

There is one feature in connection with automatic signals 
about which some railway officers feel uneasiness, and that 
has relation to the steps that have to be taken when a driver 
finds an automatic signal at danger. As a rule he must stop 
for a short period from one to four minutes and then pro- 
ceed " under caution." This, they fear, may one day lead to 
trouble, and especially if the signals be not kept in the highest 
state of perfection. If not well looked after they will be 
frequently out of order, and that will often lead to signals 
being passed at danger and drivers finding the section clear. 
This will in time lead them to treat signals so shown with 
less respect and to travel through the section at ordinary 
speed, and some day it will be found that the signal was 
"on " for its legitimate purpose, but too late to avert 
disaster. These fears arc no doubt exaggerated, but they 
indicate the state of mind of some British officers. 

The first practical automatic signals in Great Britain 
were those fixed on the Liverpool Overhead R. in 1893 and 
which are still in work. 

The stations on this line, being both numerous and close 
to each other, the estimate for the initial cost, and subse- 
quent working and maintenance, of ordinary mechanical sig- 
nalling was so large that the engineers sought for some other 
method of meeting the Board of Trade requirements at a 
less cost, with the result that the automatic system invented 
by the late Mr. I. A. Timmis, of \Yestminstcr, was adopted. 

The line consisted then of two terminal stations, two 
junctions and 15 intermediate stations. Of these 13 stations 
are protected by automatic signals, there being two signals 
for each line a home signal about 100 ft. in the rear of the 
station and a starting signal in advance acting as the distant 
for the next section. 

The line is not equipped with "Track-Circuit," but is 
controlled by electrical contacts and the signals stand nor- 
mally at " clear." Each signal is put to " danger " as the 
train passes it, and when a train enters a station it puts the 
home signal at the station to " danger " and lowers the home 
signal at the station in the rear. When it leaves a station it 
puts the starting signal there to " danger " and lowers the 
starting signal to " clear " at the station in the rear. Each 
train is, therefore, protected by two absolute stop signals. 

The locking and unlocking device is not of the usual 
treadle arrangement, but is a lever by the side of the line 
and contact is made by the lever being struck by a bar at- 
tached to the last vehicle on the train, so that the section is 
not cleared should the train have broken loose, and the whole 
of it not have passed out of the section. 

This seems a simple and economical method of signalling, 
and would no doubt suit any railway on which the class of 
traffic is similar, and where it is the practice to run the 

trains in blocks that arc not broken up, and where there is no 
shunting or attaching of vehicles. On such a line there are 
no sufficient reasons for going to the great expense of 
" Track-Circuits," and in that event automatic signalling is, 
without doubt, the most economical form of signalling that 
could be found. 

Nothing more was done with automatic signalling until 
the autumn of 1900, when a visit was paid to the United 
States by two officers of the L. and Southwestern R., Mr. 
Jacomb Hood, the chief engineer, and Mr. Sam Fay, then 
superintendent of that line and now general manager of the 
Great Central R. As a result, a power plant was laid down 
for working the points and signals at Grateley by the Low- 
Pressure Pneumatic System, and the six miles of double line 
between Grateley and Andover were equipped with automatic 
signals on the same principle and operated by power 
generated at the source of supply at Grateley. This was 
opened in the summer of 1901. 

In August, 1901, the North Eastern Co. ordered a small 
trial installation of the Hall System, but after a visit paid to 
America by Sir Geo. Gibb and some of the other leading 
officials, this small order was cancelled and one given for 
the equipment of the line from Alnc to Thirsk, a distance 
of ii miles. This work was opened in June, 1905. 

The object of the Andovcr-Gratcley installation was to 
increase the carrying capacity of the line, but in the Alne- 
Thirsk installation an attempt was made to reduce the work- 
ing expenses by withdrawing signalmen from existing 
signal-boxes. Between Alne and Thirsk (Green Lanes box) 
there were six signal-boxes, one of which was an intermediate 
block post without any points ; two worked roadside stations 
with an up and a down siding connection and a crossover 
between the main lines; two worked double line junctions 
(both with facing and trailing points and a crossover, and 
one with a siding connection) ; and one worked a roadside 
station with an up and a down siding connection and a 
crossover, but having a busier traffic than the two other 
roadside stations. 

As was observed earlier in this chapter the work of a 
signalman consists of signalling trains on the block instru- 
ments, lowering the outdoor signals for the passage of 
trains, and pulling over the point levers when any shunting 
has to be done. It follows, then, that where there is very 
little shunting to be done the greater part of the signalman's 
time is occupied in putting the trains " in block " and work- 
ing the signals. By the use of automatic signals these 
operations can be performed without the signalman, and 
he is only required when any shunting has to be done. As 
the traffic at two of the roadside stations between Alne and 
Thirsk is rather light these two signal boxes have been 
closed, and the signals that protect the station and the con- 
nections are the automatic signals in the section. The 
signal-boxes arc retained, also the locking frame, rodding, 
etc., for the signalmen to work the points when any shunt- 
ing has to be done. The intermediate signal-box has been 
removed and one of the junction signal-boxes is closed at 
night when the traffic on the branch they give access to is 



over. The carrying capacity of this section of the North 
Eastern main line has been increased, as the number of 
block sections is now 15 instead of / under the usual system 
of working. 

The success of the Grateley-Andover installation was 
sufficient to justify its extension to Woking and Basing- 
stoke. On this section of the~L. and South Western R. there 
are four lines and all the points and signals at the stations 
and junctions are worked by power on the Low-Pressure 
Pneumatic System, with automatic signals for dividing the 
intermediate lengths into sections of about 1,500 yards 

On the Great Central R. automatic signals on the Low- 
Pressure Pneumatic System have been provided for break- 
ing the long block section on the up road between Whet- 
stone and Ashby Magna, also for dividing the block section 
on the up line through Woodhead Tunnel. 

On the Great Western R. the section of the four main lines 
between Pangbourne and Goring, 2^ miles, has been equipped 
with automatic signals and " Track-Circuit," which were 
brought into use in August, 1907. Seeing that the Great 
Western is about the best-signalled railway in the country, 
this installation demonstrates the confidence now placed in 
"Track-Circuits" and automatic signals. See p. 335. 

The Lancashire and Yorkshire R. have a section of 
automatically signalled line on the Westinghouse system 
near Rochdale, and also some automatic signals on the 
electrically operated Liverpool-Southport line. 

The Great Northern and City electric tube railway has 
been provided with automatic signals by Spagnoletti and Co. 

But the most interesting installation of automatic sig- 
nalling in Great Britain is on the Metropolitan District R., 
the " Bakerloo," the " Piccadilly" and the Hampstead 
tube railways, and which was carried out by the Westing- 
house Brake Co. 

On British lines it is the custom to put all the automatic 
signals on a telephone circuit connected to the signal-boxes 
on either side. This is verv useful, as it enables engine- 
men to make enquiries from the signalmen if they be de- 
tained at a stop signal for no apparent reason, or in case 
of an accident or breakdown to communicate with them. 
The telephones are also useful to telegraph linesmen and 
signal repairmen. 

Where the block system is replaced by automatic signals, 
and there is a long stretch of line between the signal-boxes 
that are open, the entrance of trains from the block station 
in the rear is " belled " to the box in advance. As some 
time may elapse from the giving of the bell signal to the 
arrival of the train, according to its speed and the calls it 
has to make en mute, the signalman at the advance box 
may be in a difficulty in case he has a shunting movement 
to make to know the whereabouts of the approaching train. 
The Author suggests as a remedy for this that a row of 
electrical discs be provided in the signal-box, connected to 
the different sections, which should indicate which sections 
were occupied. 

Where two sections of automatic working converge at 
at a junction it is, of course, possible for trains from the 

two converging points to approach the junction simul- 
taneously, which is contrary to junction working. To meet 
this outer home signals should be provided on each of t he- 
converging lines. 

Automatic Signalling on the Continent. 

The first installation of automatic signals on the Con- 
tinent was on a privately owned railway in Belgium the 
Ghent-Wondelgem line, 5 kilos, in length since taken over 
by the Belgian State R., but the signals have been removed 
as the course of the line was changed. These were Hall 
disc signals. 

The Hall system has been installed on the P. Lyons M.R. 
between Laroche and Auxerre, a distance of 38 kilos (2 Y6 
miles). The ordinary signals of the railway are used. Also 
between Bordeaux and Langon on the C. de fer du Midi, 
a distance of 42 kilos (.>b miles). Disc signals are employed. 

The Lyons Co. are also considering the desirability of pro- 
viding automatic signals instead of intermediate block posts 
on 10 kilos of road near Nimes. 

The Chemin de fer Metropolitaine dc Paris is an electrically 
worked railway protected by automatic signals. " Track- 
Circuit " is not employed, the signals being lowered and put 
to danger by relays connected with treadles actuated by the 
deflection of the rails. That part of the line that was first 
opened was signalled by the Hall Signal Co. Originally it 
was on their " normal-clear " method, but it has now been 
altered to " normal-danger " throughout. 

Each station has two signals, one at the rear to protect 
trains standing in the station, and one at the outgoing end 
to act as a starting signal. In cases where there is consider- 
able distance between two stations an intermediate stop signal 
is provided. 

The extensions of the Metropolitaine have been signalled 
automatically on a system designed by some of their own 
officers and described in chapter XIV. 

On the Austrian Sudbahn (Southern R.) there is a trial in- 
stallation of automatic signals on a short length of line. 

The reasons why greater progress has not been made 
on the Continent may perhaps be judged by the observations 
made by Mons. Weissenbruch, one of the chief engineers 
of the Belgian State Rs., in an article* on the Siemens- 
Halske electric power signalling installation at Antwerp. 

The management has arrived at the conclusion ihat the application 
oi such a system (automatic signals) to the lines of the Belgian State 
R. is not justifiable on economic grounds. These lines have a 
total length of 4,050 kilometers (2, 53(3-8 miles), and include 1,1(15 
stations and junctions, and 3,270 protected level-crossings. It is thus 
always possible to divide the line into separate sections so that the block 
appliances are operaled by a signalman, by a pointsman or by a gate- 
keeper, and actuallv it costs very little to operate the different appliances, 
liven if that were not the case, and if a considerable number of extra 
employees were required for these operations, it would still be preferable 
to have a number of separate cabins and the operations carried out by 
hand. We can quite understand that in America automatic working has 
become absolutely necessary in certain cases, because it is very difficult 
to find signalmen who are willing to stop in cabins which are far away 
from any human habitations, and, moreover, there it is comparatively 
easier to find skilled electricians. But in Belgium just the opposite is 
the case. Here (in Belgium! there is an ample and cheap supply of 
labour, of comparatively ignorant men who have no technical know- 
ledge, but have sufficient intelligence to be able to carry out the duties 
of a signalman, whereas really skilled electricians are scarce. 

A* far as safetv is concerned, we do not in Belgium consider the 
automatic systems to be any advance on the systems operated by hand. 
It is true that there are very well constructed automatic apparatus, and 

'The Raihoay Age, 22nd April. 1900. 


I>01 V K R RAIL W AY SI G N A L L I N G . 

it would perhaps be wrong to say thai under favourable, conditions they 
do not give the same degree of safety as the block system with signals 
normally at danger, interlocked with the track appliances and operate< 
bv hand. But it cannot at present be denied that even with the most 
careful maintenance the cases of failure of automatic appliances are 
more numerous than those of the appliances operated by hand. In such 
cases it is a definite advantage that the signalman is present, and tn 
his intelligence may act as an immediate substitute for the appliance 
which has broken down. 

In cases of failure the only guarantees left, in the case of an automatic 
block system, are the regulations issued to the trainmen, and it is not 
certain that they will be obeyed. But with a non-automatic block, tl 
vigilance of the signalmen may temporarily compensate for the absence 
of the appliances; moreover," the signalmen of the different cabins 
mutually control each other, and this practically ensures the proper 
carrying out of the regulations. 

Automatic Signalling in America. 

The earliest successful application in America was in 1871, 
but no great advance was made until during 1883-5, when 
over 650 automatic signals were in operation. By the end 
of 1899 nearly 7,000 were in use, and now it is estimated 
that there are over 30,000. In January, 1901, there were 
2,294 miles of track protected by automatic signals in America 
which, according to a return in The Railway Age of May 
I7th, 1907, had been increased to 8,561 miles (2,750 single 
and 5,790 two or more lines). 

In America, nearly all automatic signalling is upon the 
" Track-Circuit System " by insulating the whole length of 
the line but there is still a quantity of the " Wire System " 
in use, whereby the locking and unlocking is done by an 
electrical treadle, without the continuous security given by a 
" Track-Circuit." The two original systems were the Hall, 
with electricity throughout, and the YYestinghousc, in which 
electrical contacts open and close valves of cylinders in which 
pistons connected to signals are worked by compressed air. 

The original signals were of disc or " banner " form, but 
objections having been raised to this class of signal, a signal 
of the standard semaphore type is now provided also. 

The objections that have been raised to the "disc-signals" 
may be briefly summarised : 

1. That the case enclosing the disc can be covered by a 
damp, sticky snow, and the signal obscured. 

2. That the glass can be broken, and the disc stem bent 
by a missile thrown at the signal. 

3. That the face of the case may reflect the sunlight at 
such an angle as to render the signal-indicator indistinct 
during a portion of the time a driver is approaching it. 

Advocates of the disc-signal retaliate by saying that the 
semaphore pattern is not perfect, and is to be objected to 
because : 

1. It may be frozen in the "clear" position by wet snow 
falling and freezing on the blade, and the connection between 
the blade and its support, which cannot happen to the en- 
closed disc. 

2. Its greater first cost and consumption of power for 
operation as compared with the disc. 

3. The greater liability of derangement of the semaphore 
as at present installed, and the consequent necessity for a 
larger and higher skilled force for maintenance. 

The Chicago and North Western RR. adhere to the disc 
signal, and use it throughout the whole of their vast system. 

Mr. Edward C. Carter, the chief engineer of that com- 
pany, presented a report on Automatic Signalling to the In- 
ternational Railway Congress held in Paris in 1900. This 
gentleman has had some considerable experience with auto- 
matic signals, and dealing with these objections, he said in 
connection with disc signals that snow on the case is similar 
to snow on the spectacle of any signal at night, and therefore 
being an imperfect signal, can only lead to delay. The disc 
showing a clear signal owing to its having been damaged 
is a mishap most unlikely to pass unnoticed by a driver, and 
the signal would have to stick at just the right point in order 
to avoid being improperly displayed, and as regards the third 
objection, there will always be a certain space through which 
the driver passes in approaching the signal, in which the 
signal can be observed without its being obscured by 

Mr. Carter also dealt with the objections to semaphore 
signals. He said that heavy counter-weighting will dispose 
of the first objection, or better still to adopt the " normal- 
danger " system whereby a signal is only " off " for a short 
time, and therefore there is every probability that if the signal 
froze at all, it would freeze in the " on " position. As to the 
greater first cost and increased cost for working, these are 
not points that anyone would consider who was satisfied that 
the semaphore was better than the disc, and lastly, the 
greater liability to derangement is a matter that is being 

Mr. Carter also gave the following figures, shewing the 
number of automatic signals in the States, which, although 
now out of date, are interesting as marking the progress 
made and the changes in the various types. 


1-1 Jy 

1 " 





i^ t/3 





























1 885 














1 f\ 





















* . 







35 '> 






















2 7 





















These figures have been greatly increased since then. 
The Hall Signal Co. inform the Author that they have 5,800 


electric disc signals in work or on order, 2,200 electro-motor 
signals and 4,283 electro-gas signals. The Union Switch 
and Signal Co. state that they have 15,700 automatic electric 
semaphores and 17,156 automatic signals of all kinds. In- 
cluded in these figures will be the semi-automatic signals 
that are worked by power-operated and mechanically-operated 
plants and controlled by the track circuit. Both the General 
Railwav Signal Co. and the General Electric Co. have auto- 
matic signals. 

From Mr. Pratt 's report to the International Railway 
Congress of 1905, sixty-two American railroads reported 
that their automatic signals made over 405 million move- 
ments in one year. Of these, 52 companies reported nearly 

twenty thousand failures out of nearly 39^ million move- 
ments. Twenty-nine companies reported 186 cases where 
signals indicated " clear " when the section was not clear, 
but what proportion of the total movements is not given. 

There is one matter connected with automatic signalling 
in America that might be improved. From what he has 
seen the Author thinks that not sufficient attention is paid 
to the location of signals in relation to the position of con- 
nections with the main line, stations and level crossings. In 
many cases he noticed that by placing an automatic signal 
in a different position it would afford better protection to 
trains waiting at those connections, standing in stations, etc. 

Fig. 2796. Top Mast Motor Signal Great Northern Railroad; t'.S.A. 
(For description see page 146.^ 

K 2 



THE general arrangements adopted in America are as shown 
on fig. 243, which represents a length of line between two 
stations and divided into three sections, A B, B C, CD, the 
section A B being protected by the home signal a, the section 
B C by the home signal c and its distant b, and the section 
C D by the home signal e and its distant d. The home 
signal g is for the protection of the station there and for 
the next section. Its distant is /, and the distant h under 
the home g is for the home signal next in advance, which 
is not shown in the sketch. 

It will be noticed that a train is proceeding from one 
station to the other and is now in section C D. In its pas- 
sage it found all the signals off, and as they were passed 
they were automatically put to danger by the train actuating 
relays at the insulated joints which mark the boundaries 
of the different sections. Starting from the station on the 
right, it reached insulated joint /' and entered the section 
from j to k, whereby the current flowing from the battery 
fixed near k was short-circuited from signal a so that the arm 

C D is not lowered until the train has left the station, as the 
releasing relay at insulated joint n is fixed in advance, and 
the -section C D includes the station, so that trains stopping 
there are efficiently protected, but when they are on their 
way and past the joint n the home signal c and the distant 
</ are pulled "off." 

In some cases possibly the majoritv the station would 
be protected by signal e and signal g would be fixed in 
advance of the station, near joint n. 

In America it is the custom on roads automatically sig- 
nalled for siding and other connections, except running junc- 
tions, to be worked by the guards and shunters from ground 
frames, with a certain amount of freedom not met with in 
the United Kingdom. These movements are, however, 
properly protected, as it is one of the principles of " Track- 
Circuits" that an open switch, as well as vehicles standing 
on the line, cause short-circuits and the signals behind a train 
engaged in shunting operations remain at danger and there- 
fore afford the same security as if a signal-box were there. In 

r ff I a ? T J Tint, 

'///////////////A *- - -Section. A. B: * SeclionB.C. * SectionC.D. i- - 1/ , 7 -^- -, 

" m ^ m _WMiUitltU(iaU{LlZA 


joint j 

Train. J 

Insulated, ,) Insulated, 

joint h joint: I. ' 

Insulated joint 
Fig. 243. Sketch of Aulomaiic Signals as arranged in America. 

Insulated / 
joint 7t 

rises to danger, and at the same time carries with it the distant 
arm b. This latter is done through the usual slotting 
arrangements, whereby a top arm controls a lower distant 
arm so that the latter cannot be " off " when the upper arm 
is at danger. On passing joint k the home signal c (and 
with it the distant d) goes to danger, and the train being 
protected by that signal, the electrical circuit to signal a is 
again completed by the battery being joined up, and the 
signal is pulled " off," the distant b being left "on " as its 
home signal c is at danger. \Yhen passing joint I the home 
signal e is thrown up, together with the distant /, and the 
train being under its protection, the home signal c is pulled 
"off" and also its distant b, the distant il remaining "on" to 
indicate the position of e. 

Matters are now as seen in fig. 243, the train being pro- 
tected by the home e and the distant d, and the home signal 
g being " off " for the train to enter the station. Imme- 
diately after passing the signal g the train passes over the 
insulated joint m, which puts to danger the signal g and the 
distant h, but the home signal e at the entrance to section 

order, however, to warn men engaged in shunting that a train 
is approaching, or to warn them of the approach of a train on 
the opposite road, it is customary to fix an electric bell at each 
set of switches that rings when a train is coming. If a bell 
is not provided an indicator is given. 

At signal-boxes at the entrance to an automatic section 
the starting or last stop signal is controlled by the " Track- 
Circuit " by replacers similar to those illustrated in chapter 
V., so that such signals cannot be lowered if the section be 
not clear, and the distant signal worked from the signal- 
box is also electrically controlled so that it cannot be lowered 
unless the starting signal is "off." 

Where the distant for a mechanically worked signal-box 
is situate as the lower arm on an automatic signal, the former 
signal is worked by wire and controlled in the usual way, 
but where an automatic distant comes on a mechanically 
worked home or starting signal, the distant is worked elec- 
trically and the slot is an electric one, so arranged that when 
the distant portion is pulled " off " and the upper arm is 
" on " the two upright rods are put in gear so that when 



the top arm is lowered both arms come "off" together. Should 
the top arm be already " off " the distant arm is pulled "off" 
electrically. The distant can be put to danger independent of 
the upper arm, but should the latter go on first both go to 
danger together. It will, however, be found better to work 
both arms like automatic signals, ever, should this necessitate 
running wire from the mechanical box, as on the North 
Eastern R. of England (see chapter XIII). 

Hall Disc Signal. 

The original Hall signals in America were of the disc- 
type, there being about 6,000 in use, but electric signals 
of the usual semaphore type are now generally used. 

Interest in the disc signal has, however, revived because 
-0, of its simplicity, few working parts, small 

cost of installation and subsequent mainten- 
ance, and the little power required to work 
the disc. An advantage also lies in the fact 
that all the moving parts are securely 
housed and cannot be affected by the 

The signal is illustrated by fig. 244. The 
drum-head may also be attached to the side 
of wooden posts or on to gantries where 
bridges of signals are used. The larger 
opening b is for the day indication and the 
smaller one a for the night. 

In fig. 245 is shown an internal back view 
of the upper part of the disc signal. It is 
4tt. bins, high and 3ft. Sins. wide. The 
case is strongly constructed, being made of 
two thicknesses of white pine panels, the 
outside ones being perpendicular and the 
inside horizontal, heavy building paper 

F'g- 244- 
Hall Disc Signal. 

being laid between, the whole preventing the splitting of 
the case. The front is covered with sheet steel. 

In the centre of the front of the signal is a circular plain 
glass, i8fins. in diameter, in front of the disc r (b in fig. 244!, 
and in the upper part of the front of the case is a second 
circular piece of plain glass 7in. diam. 

At the back is a door which is shown open in fig. 24^, 
and in this there is a circular glass /, either opal or painted 
white, 22fins. diam. with a plain piece in the centre ^ins. 
diam. Above the door is placed a signal lamp, held in 
position by the lamp bracket g to the left of the upper lens 
(a fig. 244). 

The inside mechanism consists of two discs c and d, as 
shown in fig. 246. The lower, c, is made of red cashmere 
and is I7ins. diam. It is normally suspended between the 
front plain glass and the back white glass, and gives a 
danger signal. The upper disc is of red glass, 6iins. diam. 
and is normally between the upper front glass and the signal 
lamp, and thereby gives a night indication. The two discs 
are joined together by an aluminium arm a which has a 
circular armature. This, when the signal has to be cleared, 
is attracted by the clearing coils b in the centre of the case 
so that the lower disc is turned to the right and the upper 
disc to the left, so that a clear signal is given by the red 
disc being removed and the white background being seen. 
The upper red disc is removed from in front of the lamp so 
that a clear signal is given by night. On the rod of the 
upper disc will be noticed an armature which is attracted 
by the pole-pieces of the hold-clear coils c on the left when 
the signal is cleared. These coils are de-energised when the 
signal has to be restored to danger, so that the discs go to 
normal by gravity. 

Fig. 245. Inside of Disc Signal. 

Fig. 246. Discs, Hall Signal. 

247. Disc Signal with third Indication. 



Where a green light is used for the " clear " position, a 
third disc h is provided as illustrated by fig. 247. When 
the signal is at danger, the day indication is given by disc 
c and the night indication by disc d. When the signal is 
cleared, the green disc h rises and comes 
before the lamp in place of disc d. 

Where disc signals are used for stop 
and distant on the same post the stop 

disc is on the top and the distant disc is fixed below the stop 
disc. The current for clearing- the former passes through the 
stop disc, so that the distant cannot be put to "clear" until 
the upper stop disc is cleared. 

Hall Electro-Gas Signal. 

Whilst the disc form of signal gave satisfactory results 
it became necessary to adopt an automatic semaphore signal. 

These were operated by electric motors, 
but about 1902 the Hall Signal Co. intro- 
duced their electro-gas signal, in which 
carbonic acid gas is used for moving the 

signals and electricity for controlling the valves. Egch signal 
has its own gas tank and electric batteries. There are now 
about 4,500 of these signals in use or on order in tfje United 
States, and it is this electro-gas signal that has been erected 
on the North Eastern R. of England. 

The signal complete, as modified to British patterns, is 
illustrated by fig. 248. It is built up of six principal parts : 
the lower base containing the batteries ; the upper b,ase en- 
closing the gas mechanism ; the drawn steel-pillar ; the collar 
on the top of the upper base into which the pillar is stepped 
and cemented ; the arm carrier, which also forms the sockets 
connecting the lengths of the pillar and the pinnacle. Stages 
for the use of lampmen and signal repairmen are provided, 
if required, although these are not shown in the drawing. 
The base is bolted to a concrete foundation by four anchor 

Fig. 249 is a view of the lower part of a signal wjth the 
upper door open, showing the mechanism. 

The gas is stored in a cylinder or flask fixed in a shute 
in the ground by the side of the signal, and as the shute will . 
hold two cylinders one of them can be always full. The 
cylinders are 4ft. 3ins. high and Sins. diam. outside. They 
weigh when empty 200 Ibs., and hold at a pressure of about 
900 Ibs. per sq. in. 40 Ibs. of liquefied gas, which is reduced 
by a reducing valve, fig. 250, to 40 Ibs. working pressure. 

The details of the mechanism for a two-arm post with 
upper stop and lower distant-arms is shown by fig. 251. The 
connections on the left side lead to the upper arm, which is 
supposed to be "off." The operating cylinders, which are 
arranged vertically, are rigidly attached to the signal upright 
rods and the pistons are fixed. The gas enters through the 
piston, forces the cylinder upwards, and clears the signal. 
The admission of the gas to the working cylinders is con- 
trolled by a valve which is opened and closed by the armature 
of an electro-magnet. In automatic signalling these mag- 
nets are energized in the usual way, by a local circuit con- 
trolled by the relay of the " Track-Circuit." When a signal 
has been cleared it is held in the " clear " position by the 

\-"\g. 248. Jlall Electro-Gas Signal. 

A I' TO M A '1IC SI G X A L S . 

Fig. 249. 

Hall Electro-Gas Automatic Signal. 

Fig. 250. 

'UpryhtRodsatiachtd hen 




l ; ig. zsi. Details of Mechanism, Hall Electro-Gas Signal. 



mechanism, which is described below. Suitable circuit 
closers or electric switches are provided to insure the opera- 
tion of the two signals in proper sequence, the distant to 
be cleared after the home signal has been pulled "off." 

The gas enters the reducing valve a, fig. 251, by the pipe 
b l , and is led by pipe b- to the expansion chamber c, which 
enables the gas coming from the tank to partially expand 
before going to the cylinder and thereby prevents the gas 
freezing from a too rapid expansion. If freezing should 
occur, it produces a thin, snow-like substance too fine to 
clog the movement of valve or cylinder. It is, however, a 
waste of gas. 

From the expansion chamber the gas passes to the elec- 
trically controlled valves d d. When the signal is to be 
lowered the magnet e is energised, attracting the armature 
f 1 attached to the right angled crank g l and the valve, fig. 252, is 
operated by means of the rods g- and g 3 . The ex- 
haust valve /; is attached to # 3 and is forced against its seat, 
and supply valve i is by the same movement opened and 

Fig. 252. Ball Valve, Hall Klcrtro-(ias Signal. 

allows gas to enter the cylinder through the pipe k, and to 
force the cylinder up and lower the signal. Latch I on the 
cylinder rod engages with a step on the lever in, and the 
signal is held at "clear" because the armature f on the 
lower end of lever m is being attracted by the magnet e. 
The " buffer lever " n is provided to prevent the lever m 
striking the end of the magnet when the signal goes to 
danger, and also to hold the armature f a short distance 
from the pole of the magnet while the signal is at danger, 
to prevent them freezing together in case moisture condenses 
on them in frosty weather. The casting carrying the latch I 
also carries a roller o, which engages the pawl p when the 
upward stroke is completed, and disengages it from the 
tooth on the J_ lever q which is pivoted at s by one of its 
short arms, and the other short arm is engaged with nut r 
(fig. 252) attached to the link 5. The downward movement 
of q forces the links s g 3 also down, and opens the exhaust 
valve h and closes the supply valve i, as no further power is 
required, the signal being held at clear by the latch I resting 
on the lever m. 

The stop-signal being lowered the corresponding distant 
signal may be pulled " off " and the distant arm under the 
stop signal also released. Current to these signals passes 
through the switches i 1 t 2 , the former being to the corres- 
ponding distant signal, and t z to the lower distant arm. 
When the upright-rod rises a stud raises the rod v, which 
rotates the shaft of the switch. The switch t 3 is spare. 

When the magnet e is de-energised, as by the entrance 
of a train into the section, the armature f is released and 
the lever m swings back and frees latch I. The signal then 
goes to the danger position by gravity. The cylinder acts 
as a " dash-pot " on account of the check ball-valve in (fig. 
252) partly closing the exhaust. 

The gauge on the reducing valve has two pointers and 
shows both the pressure in the supply tank and the working 

The cylinders and pistons are made of phosphor bronze 
and are ground to fit, no packing rings are used and no 
lubrication is necessary. The area of the piston is 5 sq. ins. 
By using 40 Ibs. pressure, which makes a force of 200 Ibs., 
there will be at least a margin of 50 Ibs. over the weight 
of the ordinary spectacle casting. This margin can be in- 
creased to anything desired by increasing the gas pressure. 
With this pressure and a 60 movement of the arm, 250 
signal movements are made per Ib. of gas, or 12,500 
from each 50 Ib. tank of gas. 

The number of cells of battery necessary to operate the 
magnets and the amount of electrical energy consumed are 
obviously dependant on the spacing of the signals and the 
frequency of trains. From 4 to f> cells of the " Edison " or 
" Gordon " type are used on wire circuits of average length, 
and as only 18 mil-amperes are required to hold the signal 
in the " clear " position,' the battery consumption is 
relatively slight. Assuming the price of liquefied gas at 2d. 
per Ib. , and estimating 250 signal movements per Ib. of gas, 
the cost of operating signal 1,000 times is 8d., plus the cost 
of battery consumed. Those figures are based on the as- 
sumption of perfect maintenance and have often been equalled 
in actual service. 

Hull Electro-Gas Signal, German 1'altcrn. 

Messrs. Siemens and Halske are the agents in Germany 
for the electro-gas signal, and they have modified it. 




It should be remembered that in Germany signals must 
be returned to the "on " position by force, and not allowed 
to do so by gravity alone. 

Further, the arrangement now about to be described 
applies to other classes of signals where fluid is employed 
as the operating power. Fig. 253 illustrates the modifica- 
tion. When the signal may be lowered the magnet i is 
energised and attracts armature 7 so that port 3 is closed and 
4 is opened and no air (or gas) flows from the reservoir to 
the bottom of the larger cylinder B, but it flows to the smaller 
cylinder .4 so that the lever C is turned and this lowers the 
signal. At the same time the piston in the larger cylinder B, 
being also coupled to the lever C, is forced down, the air 
escaping to the atmosphere through valve 5, which is closed 
automatically when it comes into contact with the bottom of 
the cylinder. 

The signal is restored to the "on " position by the mag- 
net i being de-energised so that the armature 7 falls away 
and closes port 4 and opens port 3 so that air that is 
in the smaller cylinder .4 passes into and expands in cylinder 
B and raises the piston and the signal without any additional 
expenditure of air or gas. 

In order to ensure that this is done the magnet 2 is 
energised and this forces the armature 7 downwards. Fur- 
ther, to prevent the signal being operated improperly, the 
catch 6 is provided which is engaged by the locking lever 
6 X when the armature 7 rises, but when not so engaged, 
should the lever C move downwards, the catch 6 would be 

drawn by its spring into the slot on the lever and there held. 
Hall Electro-Motor Automatic Signal. 

Where an all-electric motor signal is demanded the Hall 
Signal Co. recommend their type F, designed and patented 
by their electrical engineer, Mr. Clarence W. Coleman. 

The outward construction of this signal is similar to that 
of the electro-gas signal. The internal mechanism is illus- 
trated by figs. 254-256. 

A two-arm stop and distant signal is here represented. 
The stop signal is coupled to the upright rod a and it is at 
clear. The distant signal is coupled to the rod a" and it is 
" on." Each upright rod has a bracket b attached to it on 
which is pivotted a thrust piece r having a bevel c- and carry- 
ing a latch d and held in position by the spring d~. The 
driving wheel e, driven by the motor / through the gear 
wheel e", has upon it two rollers, e z e 4 , one on each side. The 
roller e 3 is on the left side and is connected with the stop 
signal, and the other, e 4 , is on the right side for the distant. 
Attached to the side of each upright rod is a latch ^ pivotted 
at g- and having a toothed projection g 3 at the top and a 
roller o- 4 at the bottom. 

Let it be assumed that, as shown in the illustrations, the 
stop signal is "off" and the lower distant arm has to be 
cleared. When the stop signal next in advance is lowered a 
current is sent through its circuit controller to the circuit 
controller h, whereby the magnet / is energised and the 
motor started, so that the driving wheel revolves and the 

Hall Electro-Motor Automatic Signal. 



roller e 4 comes under the bevel c 2 of the thrust piece c. If 
from any cause the motor were started before all was in order 
the magnet j would not be energised. Consequently the 
armature fc 2 on the long lever k would not be attracted. It 
will be seen that this lover binds against the roller g 4 . The 
effect of the attraction of the armature is to hold this roller 
and consequently the latch g and the upright rod are raised, 
as the roller e* raises the thrust piece. 

The long lever k is attached to the shaft /, which has an 
arm m with a latch m-. On the upright rod is another latch n 
with a spring to keep it in position. Just before the signal is 
cleared the latch n is pushed past the latch m"-, and then the 
moment arrives when the roller e 4 passes from under the 
thrust piece c, and the latter would fall and the signal go to 
danger were it not for the latch ; coming to rest on the latch 
n ! . But immediately the magnet j is de-energised and the 
long lever fe falls away, the latch m 2 , being on the same shaft 
as the lever *, releases the latch n and the signal goes to 

On each upright rod is a roller o, binding against which 
is the long arm p of a crank, the other end of which is coupled 
to a rod q leading to the circuit -controller, whereby current 
is cut off, and in the case of a stop signal a circuit is com- 
pleted to its sympathetic distant and its 
own lower arm. By the shape of the long 
arm /> the roller o cannot raise the crank 
to operate the circuit controller until the 
signal arm is properly "off." The arm p 
is then turned as is shown in connection 
with the stop signal on the left, and con- 
sequently when the signal rod falls the 
arm p is restored and the circuit con- 
troller again turned. 

Westinghouse or Union Switch and 
Signal System. 

This system has more than one pattern 
of automatic signal. The electric disc 
signal is similar to that seen in fig. 257; 
the illustration on the left being as seen 
by a driver and that on the right is the 
back view showing the position of the 
lamp. The upper of the other illustra- 
tions is the "off" and the lower the "on" 
position of the disc, which is attached to 
the armature of the magnet. When ener- 
gised the magnet causes the armature to 
revolve a quarter of a circle from the 
danger position (as in the lower illustra- 
tion) to the all-right position, as seen in 
the other view. The disc is made of thin 
red silk secured to a brass wire frame. 
Clockwork Signal. 

One of the earlier designed signals was 
of the open disc or banner form, in which 
clockwork was introduced. This signal 
is seen in fig. 258, the left-hand illustra- 
tion being- the signal at danger and the 
right-hand when "off." On some signals 

of this form the discs are provided with " wings " which 
present an oval shape in the opening when the signal is "off." 

The target signal seen in the upper centre part of fig. 
258 is a combination of the same class of signal. It has 
four sides, two of which are presented as seen in the left- 
hand view when the signal is " off," and the other two when 
the signal is " on," as seen in the other illustration. 

The clockwork mechanism is fixed in the upper part of 
the signal post under the disc, whilst the weight and chain 
for the clock work up and down the inside of the pillar, the 
chain being wound round the shaft in the lower part of the 
clock. The shaft is continued outside, and wound up from 
time to time as may be required. On the right is a magnet, 
the armature of which holds one of the two " flops," and 
when this is released electrically the long end falls and the 
short end raises the catch, which will be seen pivoted near 
the centre of the magnet. When this catch is raised it re- 
leases the four-arm crank and allows it to travel a quarter 
of a circle, when the pin on the next crank is caught by 
the catch. In its travel a geared wheel on the shaft turns 
the shaft (shown broken in the upper part of the illustra- 
tion), which causes the signal disc to revolve. This 
mechanism is similar to that of an ordinary escapement. 

fivlny Engutttr 

Fig. 257. Union Switch and Signal Co.'s Automatic Disc Signal. 



Fig. 258. Union Switch and Signal C'o.'s Clockwork Automatic Signal. 

The action also raises the other " flop " and puts it into 
gear for the next movement, which will allow the shaft to 
turn another quarter. The first movement will be made 
when the train passes the signal. It then actuates an elec- 
trical relay which energises the magnet, and by the means 
already described puts the disc to the danger position. After 
the train has gone the required distance another relay causes 
the magnet to be again energised, and the disc revolves to 
clear. It will be understood that, different to an all-elec- 
trical disc signal which goes up and down, the clockwork 
signal always revolves in one direction. 

When a signal is wound up it will last for 600 indications, 
viz., 300 "on " and 300 " off " positions, and arrangements 
are made whereby when the clockwork is run down the signal 
must stop at danger. 

Electro-Pneumatic Automatic Signal. 

The operation of points and signals by compressed air 

under the \Yestinghouse system, combined with the use of 

' Track-Circuits," gave the Union Switch and Signal Co. a 

new method of automatic signalling in which semaphore 

signals were employed. 

Fig. 259. Signal Motor, \Yestinghouse System. 

1 he compressed air for working these signals is conveyed 
in a pipe line from the power station, and as such a station 
can readily serve all the signals within a range of 20 miles, 
only one station is necessary for every 40 miles. 

The general arrangements are as shown in fi^. 2<~o. 

The divisions of the " Track-Circuits " are seen at a a. 
.\ -\ .\ are the batteries connected to the insulated rails and 
operating electrical relays B B B by which the current is 
switched from the batteries C C C to the electro-magnets 
E on the signals. 

Three signal-posts are shown in fig. 260, that on the 
extreme right indicating that section ahead is " clear " and 
free for the light engine shown in the diagram to enter the 
section. This engine has just passed the second signal in 
the illustration, and the current through the rails is thus 
short-circuited and the magnet of the signal cylinder de- 
energised, so that the signal goes to danger as already ex- 
plained. Connected to the upper arm is a commutator F, 
by which the lower arm and the corresponding distant on 
the stop signal in the rear is controlled. 


POWER KAIL W A V S I G N ,1 L L I N (1 . 

Fig. 260. Westinghouse Sy 

The cylinder for moving- the signal is seen in fig. 259. 
Air enters through the pipe A and passes into the passage 
B, which contains a valve C held closed by the spiral spring 
D. In the upper portion is an electro-magnet E which, 
when energised, attracts the armature F, to which is attached 
the valve C by means of the spindle G. The spindle is so 
shaped at its base as to act as an exhaust valve, and there- 
fore when the spindle is brought down it opens the passage 
H, so that air passes from A, through 
B and H, to the space above the piston J, 
which is consequently forced down the 
cylinder K. Attached to the piston is a rod 
L, which is connected to the balance lever 
on the signal. 

stem, General Arrangement. 

When the magnet E is de-energised, the armature F is 
released and is forced up by the spiral spring D, and the 
valve C is closed and the air escapes from cylinder K through 
H and M and the piston rises, assisted by the weight on the 
balance lever of the signal. 

These signals are worked on the " normal-clear " method 
whcrebv the train having entered the section in advance C D 
lowers the signal at the entrance to the rear section A B. 

Fig. 26,. 

Union Switch and Signal Co. 's 
plectric Motor Signal. 

Fier. 262. 



Upright Rod 

Fig. 264. 



Fig 266. 


Fig. 265. 

Union Switch <S~ Signal Co. 's Electric Motor Signal. 
The Union Co. 's electric motor signal is illustrated by 
fig^. 261, and the mechanism by figs. 262 to 2(16, which show 
the apparatus for a two-arm signal. The upper or " stop " 
arm (which is in the "off" position) is connected to the 
upright rod B, and the lower or "distant" arm (which is 
in the " on " position) to the upright rod, A. The working 
is more clearly shown by fig. 263, which is a vertical section 

through the nearer mechanism. 
When the motor a revolves it 
turns the geared wheels b and c, 
and the sprocket wheels d d- with 
the chain e. There are two chains, 
one for each arm, and on each 
chain is a trunnion link and roller 
/. The signal arms are attached 
by the usual upright rod (A B) to 
slot arms g g-, which swing ver- 
tically on the centre o, and the 
forked heads h h- of which rest 
on the rollers of the trunnion 
links. When the chain travels 
the slot arm is carried upwards, 
figs. 264-266, and when the full 
travel is completed the forked 
head h engages a pawl p at the 
head of the frame, so that when 
the trunnion, as it travels 
round the upper sprocket wheel, 
leaves the forked head h the 
latter is held and the signal re- 
mains in the "off" position. At 
the same time the head of the 
slot arm causes electrical con- 
tacts to be made at / whereby the 
motor circuit is automatically 
opened. A friction brake set by 
the lines of force passing out of 
the motor field soon brings the 
motor to a stand. 

Should the stop signal at the 
end of the section in advance 
have been lowered the time has 
now arrived for the lower distant 

Fig. 263 



arm to fall. The completion of the stroke of the upper arm 
having- opened the contacts at /, the magnets m arc 
energised, so that the forked head of the distant arm 
slot is held rigid. The motor is again set in motion and 
the chains e e 2 travel so that the second trunnion link and 
roller engage the forked head /t 2 and carries the distant slot 
arm similarly to the upper arm. It engages a pawl and is 
also held. 

At the other end of the slot arm of the stop signal is a 
rod fe, which on being given a downward movement operates 
a pole-changer, which opens a circuit, at Z, to the correspond- 
ing distant signal at the signal-box or signal post in the rear. 

When the signal has to be put in danger the magnets 
on the slot arm are de-energisedfigs. 264-266, which show 
the connections between the magnets m and the forked 
head h. In fig. 264 the magnets are energised and the forked 
head held. In fig. 265 they are de-energised, and the armature 
released. In fig. 266 the forked head is freed and the slot 
arm is falling by its own weight. It does not fall too 
rapidly as its action is retarded by the piston, rising in the 
cylinder n (fig. 263), acting as a pneumatic buffer. 

It may be remarked that up to the end of April, 1905, in 
addition to 6,500 electro-pneumatic automatic signals, there 
were 9,200 electro-motor automatic signals of the Union Co. 's 
type in the United States, also 1,100 of the original clock- 
work type and 356 disc or banjo signals, and these figures do 
not include signals of similar construction operated from 
signal towers. 

That these do their work well is proved by the experience 
of the Pittsburgh and Lake Erie RR. In July, 1904, that 
company took a census of their automatic signal movements, 


Fig. 268. Low Pressure Pneumatic ; Signal. 

and found the total to be 379,440. On 28 occasions trains 
were pulled up by signals being at danger for no apparent 
reason. Assuming for the sake of comparison that these 
signals should have cleared and failed to do so, the average 
of such failures is i in 18,972 movements. But only in 7 of 
these cases were the signals defective, making i defect in 
54,206 movements. In no case did a signal indicate clear 
when the line was not clear. 

n* LI u u u 1rir~ij-Tr*n 

Kig. 267. Diagram of Low-Pressure Pneumatic Signal. 

Fig. 269. Electric Low-Pressure Pneumatic Valve. 



and, rising, it causes valve e to close the lower chamber / 
and the valve e- to open the upper chamber b to supply the 
diaphragm of the signal g in fig 267. That on the right is 
for the stop signal, and when the diaphragm is raised, the 
lever h is depressed, which pulls down the rod j and lowers 
the signal. The same downward movement carries the pin 
fe in the slot /, which, when it gets to the diagonal part of the 
slot, gives a left to right movement to the slot and causes 
the contact piece m to break contact at n n and to make 
contact at o o 2 . It is through the latter contacts that the 
corresponding distant on the signal in the rear is lowered 
and the lower distant arm on the same post is freed. 

Hg. .270. Low-Pressure Pneumatic Automatic Signal. 


In Chapter XIV., where the Interborough R. of New 
York (the Subway) is described, it is recorded that in 
February, 1905, there was only i failure for every 323,594 

Low Pressure Pneumatic Signal. 

The automatic signal designed and introduced by the 
British Pneumatic Signal Co., Ltd., is illustrated by figs. 
267-271. In external appearance it resembles the others 
previously illustrated its mechanism being contained in a 
cast-iron box which forms the base of the post, which is 
either bolted to a block of concrete in the ground or to a 
bridge as shown at fig. 287 in the next chapter. 

The working of this signal is shown diagrammatically 
by fig. 267. The valves F F? which control the admission of 
the air to the diaphragms H H- are illustrated by fig. 269. 
They are worked by the electric currents of the " Track- 

Air is admitted from the main supply through the port a 
to the upper chamber b. When the magnet c is energised 
by the relay of the "Track-Circuit" the armature d is attracted Fig- 271. Details of Mechanism, Low-Pressure Pneumatic Signal. 



The operating power is air compressed to a pressure of 
about 15 Ibs. per sq. inch above the atmosphere. The main 
supply pipe A runs the whole length of the installation from 
the power house, and from which the power is taken through 
branches B and a regulating valve C to a reservoir D and 
thence through pipe E, the electro-pneumatic valves F F 2 
and the pipes G G 2 to the signal operating diaphragms H H 2 . 
The construction of the signal post, which is of iron tubes, is 
shown by fig. 270, and a detail drawing of the mechanism 
by fig- 271 and a photographic view by 268. 
Bezer's Revolving Signal. 

Fig. 272 illustrates the 
signal, with revolving 
semaphore arm, invented 
by Mr. Henry Bezer. It 
is a centrally balanced 
arm, with circuits on the 
" Normal-danger " prin- 

Its peculiar feature is 
that the arm revolves 
when the line is clear, so 
that if it revolve the 
wheels of the train may 
revolve, but if the arm 
stop the train must, or 
should, stop. 

The lamp is fixed above 
the arm, in the centre, and 
has a fixed red light, so 
that a "clear" signal is 
indicated by a flashing 

One special feature of 
this form of signal is that 
any failure is bound to 
give a danger signal, as 
motion must cease as the 
arm is centrally pivoted, 
and being heavier below 
the pivot than above it, it 
will always gravitate to a 
Fig. 272. Bezer's Revolving Signal, horizontal position. 

The following claims are made for this signal : 

1. No dependence upon automatic mechanism in any part of the 
system to change the signal from the indication of safety to that of 
danger. The presence of a train in the block beyond the signal cuts 
out the track battery, and the source of power being removed, the 
signal can only complete its last revolution. In doing so it mechanically 
forces all mechanism in the system to the required condition for the 
danger indication. As the movement of the signal indicates safety, 
should there not be a movement to the mechanism as above described, 
the signal is a danger signal because it cannot revolve. Therefore, 
failure from any cause can only indicate danger, and it is, therefore, 
impossible for this signal to remain at safety behind a train. 

2. Foreign currents (say trolley currents) cannot cause a signal to 
indicate safety behind a train. This is accomplished without extra con- 
tact points, relays or insulated rail joints. 

3. " Permissive " block is safe as regards the signalling system, 
because it is impossible for a signal to continue at safety behind a 

3. Sparking is not a source of trouble ; and should lightning weld 
any contact points, the signal can only indicate danger. 

4. In any " position " signal, the change from the indication of 
safety to that of danger is dependent upon the automatic release of the 
track relay armature. In some, following this automatic action, the 
clutch has automatically to let go. In all, following either the. first- 

Fig. 273. Three-Position Signal. 

named automatic release, or both of them, the counterweight has to 
automatically return the signal to danger. 

5. The change of the revolving signal from " safety " to " danger " 
is not dependent upon any automatic mechanism. 

The motor has only to overcome the inertia of about four pounds, 
whereas in a " position " signal it has to lift 150 pounds. 

The slight weight to gravitate the revolving signal arm to the hori- 
zontal position is sufficient for the following reason : The only inter- 
ference that can be set up to the automatic action of a weight in any 
signal is resistance in the bearings, frost, snow, etc. 

With the revolving signal any such oppositions must be overcome 
before the arm can revolve to indicate safety, and as the weight is 
called upon to act directly the arm ceases to revolve, it is not obstructed 
either by frost or tightness in bearings (these having been removed by 
the revolution of the arm), but should the weight fail to place the blade 
horizontal the signal would indicate danger, because it would not be 

In a " position " signal the weight has to act when the signal is in 
a position of rest. Therefore the weight has to overcome snow, frost, or 
resistance in bearings. 

General Electric Co.'s Signals. 

The novel feature of the signals manufactured by the 
General Electric Co., Sehenectady, U.S.A., is that the arm 
is directly driven by a motor fixed near the shaft of the 
signal arm, which dispenses with the long connections 
between the motor in the base and the signal arm. It fur- 
ther obviates the dead weight to be moved. 

In the Author's opinion there is one drawback to this 
form of signal, which is to be found in the possibility of a 
repairman not paying that amount of attention to such a 
signal on a frosty or stormy day, with a biting wind or strong 
gale. He would be inclined to hurry through his examina- 
tion, seeing that the position of the mechanism would expose 
him to the full effects of the weather, whereas, in the usual 
position at the base, if the storm were still felt there it could 
not be to the same extent, and there are possibilities of 
shelter or some protection. 

On the other hand, the position of the General Electric 
motor does ensure that if the chargeman mounts the post 
the motions of the signal will get lubrication, where, in those 
of the ordinary type, the man might content himself with 
attending to the motor batteries, etc., and chance the 



Figs. 276-277. 
Three-Position Signals. 

Fig. 274. Three-Position Signal (Back View). 

Figs. 273 and 274 show front and back views of the 
upper part of a two-arm signal. The former illustrates the 
signals used on the New York Central R., and shows the 
arms in the clear position, and in the latter both arms are 
" on " and a view is given of the mechanism. 

The General Electric Co. have a three-position signal of 
similar construction. The three positions are illustrated by 
figs. 27^-7. These represent signals on the Baltimore and 
Ohio R. 


Fig. 278. Mechanism of 3-Position Signals. 



Fig. 278 is a view of the internal mechanism of the latest 
design of signal, and shows the signal arm lowered to an 
angle of 45. 

A slot-lever a has a pawl b which is engaged by the case- 
hardened steel pin c of the driving wheel d. The slot arm is 
mounted on the spindle of the signal and it has a magnet e 
which, when energised, attracts the armature / on the other 
end of the slot lever a and holds it so that when the wheel d 
revolves owing to the motor being operated, it takes with it 
the slot lever, and consequently the signal arm spindle is 
turned and the signal lowered. 

When the arm nearly reaches the " clear " position the 
switch g makes contact with h and switches off the current 
from the motor. \t the same time the switch ;' makes con- 

Fig. 279 shows the signal and fig. 2790 is a clearer view 
of the mechanism. The largest gear wheel a is concentric 
with the signal arm spindle and it is driven by the motor b 
through the smaller gear wheels. The slot mechanism is 
carried on a disc which is rigidly attached to the arm spindle. 
There is the usual engaging lever c which, when the armature 
d? of the compound lever d is attracted by the coils e e 1 , ex- 
tends into the path of the studs /. When, therefore, the 
motor is driven the slot mechanism is turned by the wheel a, 
providing, of course, that the magnet e 3 is energised. 

It will be seen that there are two pairs of coils. The 
smaller one, e, is called the "working coil." It is of very 
low resistance less than 'i of an ohm and connected in 
scries with the motor. Its object is to give more strength to 

Fig-. 279. Top Mast 

tact, so that a low resistance circuit is closed between the 
motor brushes, which converts the motor into a powerful 
electric brake. When ihe magnet is de-energised and the 
signal has to go to normal the armature falls away and the 
pawl b gives, so that the signal goes to danger by gravity. 
At the same time the slot arm turns and its armature again 
comes in contact with the magnet ready to be again attracted 
when the signal may be lowered. 

The back is constructed so as to form secure housing for 
the mechanism. The glass cover k has been temporarily re- 
moved from the motor. The lamp is fixed on the bracket m. 

One feature of these signals is that the casing can be 
altered in position to suit the alignment of the road and is 
held in position by a set-screw. 

Top Mast Motor Signal. 

The General Railway Signal Co., of Rochester, N.Y., have 
introduced a new signal in which the motor mechanism is 
placed at the signal arm itself, and therefore upright rods are 
not required. 

Motor Signal. Fig. 2793. 

the magnet whilst the signal is being operated. The larger 
coil, e 2 , is termed the "retaining coil," and holds the signal 
in the "off" position. This coil can be of any resistance, 
but 800 ohms is recommended. When the mechanism is put 
together at the maker's factory it is so adjusted that the signal 
will be held " off " when the E.M.F. across the terminals of 
the retaining coil is 4 volts. The slot will release when the 
E.M.F. is reduced to 2 volts and the leverages are so propor- 
tioned that it will require a pull of about 4 pounds on the slot 
magnet to hold the arm "off." 

A novel feature has been introduced into this signal, which 
is the lock g which holds the rim h carried by the slot disc. 
It will be noticed from fig. 2790 that the wheel a travels some 
distance before one of the studs / engages the lever c. This 
free movement allows the stud to turn the lock g and free the 
rim h so that the arm may be lowered. The rim is so shaped 
that the lock is held away, but as soon as the slot is released 
and the arm goes to normal the weight g" 2 restores the lock 
and the arm is again fastened. 



Another novel feature is the operation of the contacts in 
the circuit breaker ;. On the end of the compound lever d is 
a roller d 3 which, when the arm is just approaching the 
"clear" position, engages the cam f* on the circuit breaker 
shaft and turns it. At the other end of the lever carrying the 
cam is a spring which reverses the shaft when the arm goes 

The dash-pot k is of the buffer type and a forged arm on 
the shaft of the slot mechanism comes against the stem of 
the dash-pot when the arm is 20 degrees from its normal 

Special care has been taken to make the apparatus acces- 
sible and so that it can readily be taken apart without special 

Fig. 279^ (p. 131) shows the signal as installed as a 
" three-position " signal on the Great Northern R. of 


Miller's Signal. 

The Author, when in America, saw in the signal shops 
of the Delaware, Lackawanna and Western R.R. at Newark, 


signal, with the signal in the normal position. The upright 
signal rod is coupled to the rod a which is attached to the 
rack b, having teeth on both sides, b" b 3 . The teeth b- en- 
gage with similar teeth on the pinion c which has an exten- 
sion carrying a cross-piece c- that forms a support for the 
driving mechanism. An escapement lock d (shown in 
detail by fig. 2800) is mounted on a pivot on the cross-piece 
c- and held freely by the clock-spring d". The escapement 
lock carries a projection d 3 which is engaged by two pins 
e- e 3 on the wheel e driven by the worm coup'ed to a motor 
of the usual" kind and in the customary manner. On the 
pinion c is mounted a magnet / which controls the armature 
g mounted loosely on the pin g- which also carries a dog g 3 
moving independently of the armature that engages in 
a notch d* in the escapement lock. Consequently, when 
the magnet / is energised and the armature g attracted, the 
dog g 3 engages in the notch d* so that the escapement lock 
is held and therefore when the motor is operated the wheel 
c revolves from left to right and the pin e- is brought into 

Figs. 280-2801:. Millar's Automatic Signals ; details of Mechanism. 

N.J., the mechanism of an automatic signal designed by Mr. 
John Millar, of Kearney, N.J. 

The inventor claims for his system that he meets a diffi- 
culty associated with automatic signals which is that the 
armature sometimes refuses to move away from the electro- 
magnet when the latter is de-energised, so that the signal 
is not free to go to normal by gravity. This may be due to 
frost or to the armature being magnetised. Further, that 
the signal is put to danger by force in case it has failed to 
be restored by gravity. 

The system is arranged for a one-arm two- or three- 
position, or a two-arm two-position signal, but it will suffice 
to describe the arrangements for a three-position signal. 

Figs. 280-2800 show the mechanism in the base of the 

contact with the projection d 3 on the escapement lock, so that 
the pinion c is turned and its teeth engage Avith those, b-, 
on the rack, and the signal rod so raised. To prevent the 
undue movement of the escapement lock, the latter is pro- 
vided with a stop </ 5 engaging with a pin c 3 on the extension 
of the pinion c. 

\Yhen the wheel e has travelled through a quarter circle 
and the signal arm has been lowered to the "caution " or 
second position, the projection 7i enters the notched standard 
Jr. This projection is shown in fig. 2800. It is carried on 
the escapement lock, and is kept in position by a clock- 
spring. It should have been stated that the pinion c is 
mounted eccentric in relation to the wheel e, so that when 
this point is reached the pin e- passes the projection d 3 , but 



the latter is engaged by pin e 3 , when the wheel is further 
driven, so that further movement is given to the pinion c 
and the signal arm lowered to the third or all-right position. 
The projection h now enters the notched standard h 3 . 

When the signal has to be restored to danger the magnet 
/ is de-energised and the armature g falls away so that the 
escapement lock is no longer held by the dog g 3 but is held 
loosely by the spring d 2 . The signal can now go to danger 
by gravity and the escapement lock, turning on the pivot on 
the cross-piece c 2 , will take the projection h out of the notched 
standard h 3 and allow the teeth b 2 on the rack -to return the 
pinion c back to normal. In fig. 280/7 is illustrated the driving 
mechanism as it is when the magnet has been de-energised 
and the signal is going to danger. 

The armature g is brought into contiguity with the mag- 
net / by the following means : On the pivoted end of the 
armature are two jaws g* g* on one of which is the friction 
roller g b . These engage with the intermediate member ;', 

Fig. 281. 

pivoted at f and having an arm y 3 . Carried in the frame- 
work is a pushing pin y 4 which, when the wheel revolves and 
the pin e 2 is being brought back to again operate the signal, 
is pushed in by the pin e 2 so that the arm y' 3 is turned, and 
this gives the necessary movement through the member y 
to the jaws g* g* and consequently the armature g is brought 
close to the magnet / ready to be attracted to the latter if 
energised, before the pin e 2 reaches the projection d 3 . If the 
magnet be not energised owing to the section ahead not 
being clear or in the case of a distant signal because the 
upper arm is not cleared, or from any other cause that re- 
quires the signal to be kept in the "on " position, the pro- 
jection d 3 will not be held rigidly and will give before the pin 
e 2 , and consequently the signal will not be lowered. 

In order to distribute the strain and not to put all the 
weight on the pin - 2 , another lock is employed, fig. 2800. 
The arm k, coupled to the dog g 3 , has a roller fe 2 resting 
on the cam y' 3 of the intermediate member y, and these form 
additional bearings for the strain of the driving mechanism. 

The usual relays are employed for energising and de- 
energising the magnet, and on the signal-rods are switches 
for joining up the current to the motor. These are so 
arranged that should the armature not have been released 
from the magnet as suggested at the opening of this 

description and the signal remains in the clear position, 
the presence of a train in the section protected by the signal 
would cause the motor to be again operated and the wheel 
to further revolve, so that the pin e 3 would come against the 
lever I- coupled to the pinion I so that the latter is turned 
and its teeth coming in contact with teeth h- brings down 
the signal-rod and puts the signal to the " on " position, 
and as soon as the signal-rod falls to the normal position the 
motor is switched out. 

The signal-rod is coupled to the piston-rod m 2 in the 
cylinder m, which acts as a dash pot as the signal goes to 

All the parts are detachable, so that any may be readily 


Three-Position Signals. 

Mr. Grafton, signal engineer of the Pennsylvania lines 
West of Pittsburg, was the first to introduce a " three- 
position " automatic signal whereby the necessity for distant 

Fig. 282. Three-Position Signals. 

signals was obviated, and this type of signal has been 
adopted on the Pennsylvania lines West of Pittsburg ; the 
Baltimore and Ohio; the Pere Marquette and the Pittsburg, 
Fort Wayne and Chicago RR. They are also in use on the 
central division of the Pennsylvania RR. See figs. 284-285/>. 

The idea is that a signal when lowered shall indicate by 
its position whether the next succeeding section be clear or 
not. The signal will not come " off " at all if the section 
immediately in advance be not clear, but if it be, the signal 
will drop to an angle of 45, and if the next section is also 
clear, then the arm will fall to an angle of 90. This does 
away with the necessity of providing distant signals. 

On the B. and O. RR. signals that are worked from 
signal-boxes (and which must not, therefore, be passed in 
the danger position, as may an automatic signal), are pro- 
vided with two arms similar to a splitting signal, and it is 
an instruction that two-arm signals must not be passed at 
danger. This, of course, means an extra arm and lamp, but 
on the other hand the provision of 'three-position signals 
renders distant arms (and lamps) unnecessary. 

Fig. 281 is a diagram of the electrical connections used 
for three-position signals by the General Electric Co. It 
will be seen that the controller is moved to positions corres- 
ponding with the occupancy of the sections. 



A bridge of automatic signals fixed by the Union Switch 
and Signal Co.' is illustrated in fig. 28.2. Three-position 
signals are employed here, the first and third arms from 
the left being in the second position, whilst the second arm 
is fully " down " and signifies that the two sections imme- 
diately in advance are " clear." 

The use of such signals would not be possible in Great 
Britain without the adoption of a third colour. In America 
red is universally used to indicate danger. The more pro- 
gressive companies have adopted green for " clear " and 
yellow for " caution " (e.g., as in the " on " position of 
distant signals), so that in three-position signals red is used 
for the first or stop position, yellow for the second or 
"caution," and green for the third or "clear" position. 

I'ig. 283. American Railway Signal Co. 's Automatic Sign;-!. 

The Author sees two objections to the use of " three- 
position " signals. The first is that a driver when running 
does not have his attention drawn so firmly to the position 
ot a signal if it be " off." All railway men will acknowledge 
that when running their attention is not attracted by signals 
in the " off " position, but, when one is sighted at danger, 
the fact is not only impressed on the mind, but to take some 
action is a natural consequence. If, therefore, a driver see 
a signal " off " and be thereby given permission to continue 
moving, he will often rest content and overlook the qualifica- 
tion of the permission, or he may possibly forget whilst 
running through the section whether the arm were in the 
second or third position. 

A second objection is that should a signal be in the third 
position and the second section should be subsequently 
fouled, the armature of the signal would be released and the 
signal would go fully to danger instead of to the second 

American Railway Signal Co.'s Signal. 

The American Railway Signal Co., of Cleveland, O. , 
introduced, in 1905, the motor driven signal illustrated by fig. 

As the Author has not had an opportunity of inspecting 
the signal, he is not in a position to deal with it fully. 

The signal has to be pulled off and not pushed off as in 
other signals. It also has a lock near the signal arm that 
prevents the arm from being lowered except by the opera- 
tion of the mechanism. 

Changes in American Signals. 

Towards the latter end of the year 1905 a proposal was 
made in the American railway world that signal arms should 
be upwardly inclined, and some such signals have been in- 
stalled on the central division of the Pennsylvania RR. 

Such signals have been in use in Germany for years, and 
more lately in Belgium. Their progress in America will be 
watched with interest. 

These signals on the Pennsylvania RR. are novel in three 
respects, and it is of interest to note that the new Union Ter- 
minus of the Pennsylvania RR. and the Baltimore and Ohio 
RR. in \\ashington, D.C., has been equipped with similar 

The principles of the new form of signalling are : the 
upwardly-inclined arm ; the three-position signal ; the stag- 
gering of the lamps of the automatic signals -which may be 
passed in the " on " position " under caution " to distin- 
guish them from absolute stop signals. 

Upwardly Inclined Arm. 

This, without doubt, is the ideal method, as in case of a 
failure it may be assumed that the signal arm will fall to 
the " on " position, whereas in England and elsewhere an 
arm, particularly if insufficiently balanced, would fall and 

Fig. 285. Three-Position Signul-, Pennsylvania R. 



Q-Creca O-White 



9 1 

O- 1 






to ni-x 

on '. 

eeei Proc 
caution at higb 
; signal on hi 
ligh speed 

eed Pn 
speed with 
gh to ne 
line. on i 


xt signal 
ed line. 


:eed at 
2ed on 
jd line. 

High and Moderate Speed. 




Proceed Proceed 

with caution on at low speed 

low speed line. on low speed line. 

Lou- Speed. On ll'ujh and Low Posts. 
Class ' A " Stop and Stay Signals for Interlocking and Train order. 

Proceed at speed, 


high speed signal at 
caution or clear. 

Proceed at 

moderate speed, next 

moderate ppeed signal 

at caution or clear. 

Stop Proceed 

and proceed to next signal 

after prepared 

waiting time. to stop. 

Class " B " Stop, wait time and proceed signals for Automatic System (also for 
distant signals wlun the latter are used as independent indication). 

Fig. 284. New Signal System, Central Division, Pennsylvania RR. 

give a " clear" signal. By the use of heavy spectacles this 
particular fear has been removed, but there still remains the 
possible evil effects of snow settling on the arm. 

This danger is the greater in America, and this is one of 
the principal reasons for the change, although the question 
of the abolition of counter-weights also assisted. Further, 
there is the advantage that the signal rods on the post may 
be of less strength, as they pull the rod to clear and not push 
it. (The rods used in this initial installation are ^in. diam. 
instead of fin.) 

The Author contends, however, that the upwardly-in- 
clined arm does not give complete immunity against trouble 
from snow, as it presents a face down which sleet and mois- 
ture will run into the bearings, and the top arm no longer 
has the shelter of the post to protect it. He finds it difficult 
to understand why the centrally balanced arm, as first used 
on our G. Northern R., has not been adopted on American 
railroads, as it would have met the snow trouble, and re- 
tained the downward inclination and so avoided a costly sig- 
nal revolution. The centrally pivotted arm also makes a good 
three-position signal. 

Th ree-Position Signal. 

The advantages of and objections to this signal have 
already been noticed. One of the advantages was the 
economy arising from no distant arm. But now all the 
advantages are destroyed, because, in future, such signals 
are to have two arms, as will be noticed later. 

" Staggering " the Lamps of Automatic Signals. 
This, in the Author's opinion, is the best feature of the 
new method. It is generally known that it has been found 
necessary to institute stringent rules for a train standing 
indefinitely at an Automatic Signal which cannot be cleared 
because it is out of order. Trains are therefore allowed to 

Fig. 2850. Three-Position Upwardly-inclined Arms and Lamps " Staggered," Fig. 2856. 

Pennsylvania RR, 



pass such a signal at danger, after coming to a stand, pro- 
viding they proceed "with caution." They are not, how- 
ever, permitted to pass a non-automatic signal at danger. 

The difficulty hitherto has been to make a clear day and 
night distinction between the two types. This has now been 
accomplished by " staggering " the lamps of automatic sig- 
nals and placing the lower lamp some 2ft. to the left of the 
upper one. 

This does not, however, provide a solution of the diffi- 
culty, referred to on p. 125, which exists on the North 
Eastern R. and the L. and South Western R., where some 
signals are worked from signal-boxes which become purely 
automatic when the boxes are closed as at nights and on 
Sundays. It is possible for a driver to be pulled up at such a 
signal at a time when he does not know whether the box be 
open or not, and consequently whether or no the signal is 
purely automatic. If the signal-box be closed then he may 
pass the signal " under caution," but if the signal-box be 
still open then he must not pass it until lowered. 

Having noted the foregoing points the diagram fig. 284 
will be readily understood. 

In the Author's opinion this scheme, instead of assisting 
to simplify signals and to make them so that they are 
quickly readable, makes them as confused as they possibly 
can be. Imagine a driver running at a high rate of speed 
having to spell such a variety of signs. Twelve different 
meanings ! ! 

Fig. 285 illustrates the upper arm in the second position, 

which, on reference to the diagram in fig. 284, it \\ill 
be found indicates " Proceed-ivith-caution-to-next-signal-oii- 
high-s peed-line." 

Fig. 2850 has three arms, the lowest arm signifying that 
there is a low-speed line. Again referring to the diagram 
and this may be done more leisurely than the unfortunate 
express driver would be able to do it will be found that the 
road has been set for a moderate speed and that the train 
may proceed at moderate speed. Fig. 285/7 shows two auto- 
matic signals fixed on a gantry. (Note the battery -box neatly 
housed in the legs of the bridge.) The " staggered " lamps 
will be observed. The signal on the right applies to the right 
hand line. (Trains in America run on the right and the arms 
point to the right.) This signal, with its upper arm " off " 
to its full extent, signifies " proceed-at-speed. Next-high- 
speed-signal-at-caittion-or-cletir." The signal on the left 
for trains in the opposite direction says " proceed-to-ue.\t- 
signal,-prepared-to-stop. " 

\Yhen revolutionising signals 'one would assume that an 
indication would be given a driver as to whether he was 
going to the right or the left at a junction, whereas all he is 
told is to go on high-speed, moderate speed or low speed 
lines, but whether to the right or left or straight-ahead he is 
not shown. There must be in America, as in Great Britain, 
many cases where space, or drivers' view, only allows and 
often barely allows for a one-arm signal. How, then, can 
a two-arm be fixed? But these are apparently brushed aside 
bv Americans as details, and below notice. 


Low-Pressure Pneumatic System. 

THE first installation of automatic signals on a steam- 
worked railway was laid down upon the Low-Pressure Pneu- 
matic System in 1901 on the L. and South Western main 
line between Grateley and Andover a length of six miles 
by the British Pneumatic Signal Co., Ltd., of Westminster. 
The installation was severely tried and found to be quite 
satisfactory, and was finally brought into ordinary service 
in April, 1902. 

The signal is described and fully illustrated by figs. 267- 
271, pp. 142-14.1. 

The salient features of the system are the low pressure 
about 15 Ibs. per sq. in. above the atmosphere of the 
operating power and the diaphragms which work the signals. 

The main power supply pipe runs the whole length of 
the installation from the power house, which is at Grateley. 
Branch pipes lead from the supply pipe to the signal 
diaphragms. The control valves are opened and closed by 
relays connected with the "Track-circuits." The sections 
are one mile long ; the distant signal applicable to each stop 
signal is on the stop signal immediately in the rear, and the 
signals are normally "clear." There is no "overlap" 
each section standing by itself. 

The arrangements of the wiring for this system are shown 
in the diagram, fig. 286, which shows three sections. 
Signal i protects section i, signal 2 section 2; signal 3 sec- 
tion 3 and signal 4 is for the next section. 

Each section is provided with a track battery B of two 
cells in multiple, connected in circuit with the rails of the 
section. Each section is isolated by insulated joints from 
those on each side of it. The track battery B is placed at 
the leaving end of the section and the relay R at the entrance. 

Each section stands alone, and therefore sections i and 
2 being unoccupied, the stop arms on signals i and 2 are 


at " clear," also the distant arm on signal i and which is ap 
plicable to the stop arm on signal 2. (The distant arm 
applicable to the stop arm on signal I is not shown.) 

The signals being "normal clear," when the train 
represented at X in section 3 entered that section, the stop 
arm on signal 3 was restored to danger behind it, and the 
stop 'arm on signal 2 was lowered. 

The entrance of the train into section 3 caused the track 
battery B 3 to be short circuited by the wheels of the train so 
that relay R3 was de-energised and armature A3 fell. As a 
consequence the circuit A3, M6, wire a, battery B8, was 
broken, and the magnet M6, associated with the stop arm 
on signal 3, de-energised. This cuts off the air supply for 
the signal and opens the exhaust so that the stop arm goes 
to danger. 

When the upper arm rises the magnet MS of the lower 
distant arm is de-energised and that signal goes to danger. 

The distant arm on signal 2 applicable to the stop arm 
on signal 3 went to the " on " position when its upper arm 
was thrown up, but in order to hold it at danger as long 
as its stop arm on signal 3 is " on," the operation of put- 
ting the stop arm on signal 3 to danger opens the circuit 
breaker C6, so interrupting the circuit B8, A3, C6, C3 (on 
signal 2), M3- The circuit was already broken at A3 by the 
train when it passed into the section, de-energising the relay 
R^, but this ensures that the distant arm corresponding to 
a stop signal is at danger as long as its stop signal is. 

When the train passes out of the section the "Track- 
Circuit " is again continuous and the current flows from 
the battery 63 through the relay R3, energising it and at- 
tracting armature A3 and again completing the circuit B8, 
A3, M6, a. This energises magnet M6 and opens the valve 
of the stop arm on signal 3 and lowers it. This closes the 
circuit breaker C6 and the current flows to the distant arm on 

section 3 

Section 2 

Section i 


i~ M 6 


Fig. 286. Wiring Diagram, Low Pressure Pneumatic Automatic Signals. 



Fig. 287. Bridge of Low Pressure Pneumat'c 

signal 2 and that arm again falls and corresponds with its 
stop arm on the signal in its rear. 

The reliability of the Low-Pressure automatic system 
having been fully demonstrated, it has been installed on the 
24 miles of four-line railway from \Yoking to Basingstoke. 
In this case the sections are about 1,500 yards long, but in 
its main features it does not differ materially from the initial 
Grateley-Andover installation. The signals are carried on 
bridges spanning the railway, as illustrated by fig. 287. 

In Chapter XX. is described the method by which those 
signals that are worked by low-pressure power plants are 
restored automatically and are thus "semi-automatic." In 
the same chapter is a description of an interesting feature 
which enables power-worked boxes to be closed at times and 
yet retain the use of the running signals as automatic signals. 

There is a length of 3 miles 416 yards between the two 
signal-boxes Woodhead East and Dunford Bridge No. i 
that control the \Yoodhead Tunnel on the main line of the 
Great Central R., and, as the up line rises on a gradient of 
i in 200, considerable difficulty was experienced in getting 
the traffic, which is very heavy, over the line, but this diffi- 

Automatic Signals, L. and >outh Western Railway. 

culty has been overcome by the provision of automatic 
signals in the tunnel, as shown by fig. 288. At Woodhead 
the up and down lines are in separate tunnels, which are 
nearly 3 miles long. The new signals in the tunnel are not 
intended for passenger trains. " Track-circuits " are pro- 
\ idcd from Dunford Bridge starting signals H G to \Yood- 
head East starting signals A B. Signal A applies to goods 
trains only and cannot be lowered unless the line be clear to 
^51 yards past the automatic home signal D. Signal B 
cannot be lowered unless the line be clear to signals G H, 
and signals D C are " off." B applies to all passenger trains 
and is lowered for goods trains also when there is a clear 
i - .id through the tunnel. 

The tunnel is very wet, and it has been thought advisable 
to limit the length of the " Track-circuit " sections to about 
440 yards. The results of this installation in such a tunnel 
will be watched with interest. 

Signals C D are equipped with full-sized spectacles, but 
signal-arms are not necessary. 

The slots on signals A B are of O'Donnell's rotary type, 
previously illustrated (fig. 125) and described. 


Hwtf flood otn . K / 23*5 
GafcritStorferj B Manhole S 

f"lfl i~' r ' 7Cie ^ TV""*! 

Automate Distant 

Cartrvlled Automatk 
Home Siqnol frnm 

Dunfara Cat/in % Dunfrrd Cobin 

Mechanical Distant Mechanical ttonm 

r) \ 

nbao*raaStn. noodhead Cab,* Slarti Synals pmiatd 

Cabin rrith Rotary Slots on Arm fto'indit All dn 

Fig. 288. Low Pressure Pneumatic Automatic Signals in Wocdhead Tunnel ; Great Central Railway. 



Audible signals are fixed 200 yards in the rear of signals 
C D to warn drivers that they are approaching the signals. 
Signal D is controlled from Dunford Bridge box, and the 
positions of signals C D are repeated in both Woodhead and 
Dunford Bridge boxes. The state of the section between 
signals A B and 531 yards on the Dunford side of signal D is 
also indicated in both the boxes. 

The Low Pressure System has also been installed on the 
Guide Bridge Widening, G.C.R., and the Clapham Junction 
Widening, L. and S.W.R., but no purely automatic signals 
are employed there. All the running signals are semi-auto- 
matic (i.e., they are restored to danger independently of the 
signalman) as described in Chapter XX. 

Between Whetstone and Ashby Magna, on the Great 
Central R., immediately south of Leicester, there is a length 
of about five miles which has been found to be too long a 
block section, particularly on the up line, and has therefore 
been provided with an automatic distant and home signal 
in the middle. 

The home signal is controlled by a section of "Track- 
Circuit " between it and Ashby Magna home signal. One 
signal is of the Hall electro-gas type and the other is a Hall 
electro-motor signal. 

The home signal, and consequently the distant signal, is 
controlled by a switch from Ashby Magna box, so that the 
signalman there can keep a train back in case he is doing any 

Hall System. 

The installation of Hall automatic signals on the North 
Eastern R. is interesting, as by it an attempt is made to 
effect economies by reducing the number of signalmen em- 
ployed. Between Alne and Thirsk (Green Lanes), N.E.R., 
there were six signal-boxes, of which one was an inter- 
mediate block post without any points ; two work roadside 
stations with an up and a down siding connection and a 
cross-over ; two work double line junctions (both with facing 
and trailing points and a cross-over, and one with a siding 
connection with one line), and one works a roadside station 
with an up and down siding connection and a cross-over, 
but the traffic is busier there than at the two other stations. 

As the traffic at the roadside stations is rather light, the 
signals there are worked automatically, the signal-box, lock- 
ing frame, rodding, etc., being retained for the signalman 
to work the points when any shunting has to be done. The 
intermediate signal-box has been closed entirely, and the two 
junction boxes are closed at night after traffic on the branch 
lines is over. 

Each line is divided into 15 sections, each of about 1,200 
yards length. Each section is protected by a stop signal 
with a corresponding distant arm on the next stop signal 
in the rear. All levers working points are preceded by a 
releasing lever which is provided with an electric lock, which 
is freed if the two stop signals in the rear be at danger. If 
either or both of them are " off " the lock is held in the 
releasing lever, which cannot therefore be pulled. In con- 
sequence, no points can be moved when an approaching 
train is within two sections of the place where shunting has 

to be done. The current to the signals passes through the 
lock on the releasing lever, so that when the lever is moved, 
the signals are kept at danger as the power supply is cut off. 

The signals are operated on the " Normal-danger " sys- 
tem and there is an overlap of 400 yards, so that before a 
stop signal can be lowered the section ahead must be clear, 
also the first 400 yards of the next section. 

Automatic signalling not only then tends to greater safety 
in working and also to economies in traffic charges, but the 
carrying capacity of a line is increased. In the Alne-Thirsk 
length the original method of working consisted of seven 
sections between Alne Station and Thirsk (Green Lanes 
box). But under the automatic signalling scheme there are 
fifteen sections. There are four lines of way south of Alne 
and north of Green Lanes, but only two lines between those 

Fig. 289 is a diagram of the line and the signalling 

The signal boxes concerned are Alne, Raskelf, Bishop- 
house Junction, Sessay Wood Junction, Pilmoor Junction, 
Sessay, Codbeck, and Green Lanes. Bishophouse and 
Sessay Wood are double line junctions, forming the base of 
a triangle leading to Sunbeck Junction on the Gilling Branch. 
Pilmoor, in addition to being a station, has a trailing con- 
nection on the down line with the Boroughbridge Branch. 
Raskelf and Sessay are country stations, and Codbeck is an 
intermediate block signal box. 

Codbeck box has been taken away, whilst Raskelf and 
Sessay are the roadside stations where the boxes are closed. 
The Gilling branch is closed at night and Bishophouse was 
only closed during the night, but Sessay Wood was always 
open, so as to act as a block post at all times. Both these 
boxes continue to be open as before. Pilmoor is closed from 
midnight to 6 a.m. 

It will thus be seen that it is possible to effect considerable 
economies in signalmen's wages, which will, however, have 
to have placed against them the maintenance of the addi- 
tional signals, interest on capital outlay, and a slightly 
higher rate of wages paid to those of the station staff who 

will act as pointsmen at the switched out boxes. 

On the down line (from York) signal i is the starting 
signal for Alne Station. It is not an automatic signal, but 
owing to there being a bad view from the signal box of the 
line northwards a section of " Track Circuit " is laid in 
between signal i and signal 2. The latter is Alne advance 
starting signal. It, too, is not an automatic signal. Both 
signals i 2 are controlled by the " Track Circuit " for the 
section immediately in front of them. By this is meant 
that when the signalman takes hold of the lever working 
signal i, a current is set up which flows through the section 
from signal 2, and if the line be clear a lock is taken out of 
the lever in the locking frame and the signal can be lowered. 
If the line be blocked, then the signal cannot be taken off. 
In the case of signal 2, the " Track Circuit " extends up to 
the end of the overlap of 400 yards past signal 4. In both 
cases the signals are put automatically to danger by the pas- 
sage of a train. 



Signal 3 is the automatic distant for signal 4, which is an 
automatic stop signal, as are signals 6 8 10, whilst 579 
are automatic distants. Signal 10 carries the splitting dis- 
tant signals n 12 for Bishophouse junction. 

There is an interesting arrangement at Bishophouse. 
The distant and home signals for travelling on the main line 
are automatic, and the signalman does not need to touch 
them for main line trains. Electric locks are fixed to the 
facing points in the down main line, so that when the 
automatic signals are " off " for an approaching train the 
facing points are locked, and vice versa when the facing 
points are moved, i.e., when the facing point bolt is with- 
drawn, the automatic signals are locked at danger. Signal 
18 is worked mechanically from the box in the usual way, 
but the distant signal 12 is lowered by power like an auto- 
matic signal, by the pulling over of the distant lever in the 

Boroughbridge Branch. Distants 30 32 apply to automatic 
home 33. 

Between Pilmoor and Green Lanes on the down line there 
is nothing of note until stop signal 45 is reached. This is an 
automatic signal and carries the splitting distants 46 47 for 
the home signals 48 49 for Green Lanes box. The home 
signals are worked mechanically, and are independent of the 
' Track Circuit " and automatic section which end here. 
The distant arms are worked by power. 

On the up-line the " Track Circuit " commences at signal 
50, which is the up starting signal for Green Lanes box. 
This is a mechanical signal, and controlled by the " Track- 
Circuit " up to the overlap past signal 52 similar to signals 
i 2 at Alne. Distant 51 is the first automatic signal. 

Between Green Lanes and Pilmoor there is nothing of 
note until automatic stop 66 is reached. This carries the 

------- >H ---- 1045 --- 

From York 

L I47A 

- 1050- 



1070 (* 1015 


I ."Ml 1 


21, 24 

:-j !-' 

Signal Bo* .. :, 

-1239 ->{< IZ6I 

--4t-- 278 *j*"~ 1074 ' 


-SSgZ^UiJ' ,29_ ,| 34-f= 

k^ .-:-:?aT~T 


iX] .^ '^v^'NJ-' I 

S<gnal Bo* jf*^ ^^ ' 


Lengths between Signals given in Yards. 

^857 -r/^~ ~^\f - 4*-- --II3I --*]< /(779 

/<360 >) 1070 *) 1270 (- *)*-- 


7I76 I 

' .-: :^ 6 - 

Sqnalto* OS'"! 





/^/O ->f /^90-- 

/^^5 ->j 

\ *, -' 



64 65 | 62 63 | 
>H ,73^ 4 

56 J^| 54 55 1 52 S3 

//74 -*t 1186 --->k- /?? *!<- 1211 -4<- //50 -4<- 1029 

Fig. 289. Hall Automatic Signals, North Eastern Railway. 


locking frame, which completes an electrical circuit sent by 
line wire to t the signal. This is done as the distant signal is 
fixed over 1,600 yards from the signal box. 

Distant 19 is the distant for the main line home-signal 
24 at Sessay Wood, and distant 20 is worked from Sunbeck 
box on the branch. 

The arrangements at Sessay \Yood are similar to those 
at Bishophouse, viz., that the main line signals (19 24 for 
the down line) are worked automatically, 22 is worked by 
power, whilst 23 from the branch is worked mechanically. 
Signal 21 is the home signal for Sunbeck. Distants 25 26 
are applicable to inner and outer homes 27 29. In the case 
of distant 25 there is no pulling off relay, being on the 
branch, so the circuit is completed on the lowering of the 
upper arm 23 by the signalman at Sessay Wood. 

At Pilmoor there are outer (27) and inner (29) home 
signals for the down main line. These are worked auto- 
matically. Signal 31 is the mechanical home signal for the 

splitting distant signals 67 68 for Sessay Wood. At Sessay 
Wood the outer (69) and inner (73) up home signals, as well 
as the distant 67, are worked automatically, similarly to 
those on the down line. The home signals 70 74 for the 
branch are worked mechanically, while the distant 68 is 

Distants 72 76 are worked from Sunbeck box. 

At Bishophouse the up main line signals (home 81 and 
distants 71 75) are worked automatically, similarly to those 
on the down line, the branch home 79 is worked mechanically, 
and distant 78 is worked by power. No. 77 is the home 
signal at Sunbeck. 

Distants 80 82 apply to the automatic home 83. 

Then there is nothing on the up line noteworthy until 
automatic stop-signal 89 is reached. This carries the split- 
ting distants 90 91 for Alne, which are also power worked. 
Signal 92 is the home signal at Alne, which is at the end of 
the automatic section and independent of the " Track- 



It will be noticed that at each station a stop signal is 
provided, which protects the station and connections. 

An interesting feature is the large number of mechanical 
signals that are actuated by power. They are provided with 
power, like automatic signals, and a line-wire is run from the 
signal box to the signal, and on the lever in the locking 
frame is a contact piece which, on the lever being pulled 
fully over, completes a circuit to the signal, which allows it 
to come " off." Directly the first movement is made to put 
the lever back the contact is broken and the signal goes to 
danger. Signals can thus be worked at any distance from a 
signal box, the accurate angle in the " off " position is as- 
sured, the certain return to danger is guaranteed, there are 
no weights for the signalman to lift, no heavy lever to drag 
over, and no signal wires that require adjustment. 

The signals so worked on this installation are the down 
(from York) distants (2) at Alne, down branch distant 
Hishophousc, down distants (2) at Green Lanes, up (from 
Thirsk) distants (2) Green Lanes, distant to branch Sessay 
Wood, the four distant signals worked from Sunbeck, also 
the two distants under the up home signals at Sunbeck, and 
the up distants (2) at Alne. 

the section as at L, and to the far end of the first " Track- 
Circuit " section is connected in the same manner the track 
battery T 1 . Normally the current from this battery flows 
out on wire I, rail 2, wire 3, through coils of electro-magnet 
' B, wire 4, rail 5, wire 6, to opposite pole of battery, thus 
energising B and holding contact points i and 2 closed. 
Relay B at signal 10 shows the normal position of the con- 
tact points on the track relay. C 1 is a " pulling-off " or cir- 
cuit-closing relay. The contact point is normally open, 
breaking the signal line circuit and causing the signal to 
stand normally in the danger position. 

The current from track battery T flows through the coils 
of relay C 1 , but coils of relay B being in the circuit there is 
not sufficient strength of current to energise relay C 1 . 
Therefore the contact point is normally open, but directly a 
train or engine say train No. 3 enters the section the 
two rails are connected by the wheels and axles making a 
path of practically no resistance from rail to rail. By cutting 
out the " Track-Circuit " the resistance of B 3 increases the 
current of battery T 4 through electro-magnet C 3 sufficient 
to close the contact. In like manner train No. 5 holds con- 
tact i on relay C 1 closed, but as No. 3 train is still on the 


i 10 

""J""*^ M 




'3 * 





\ '' 

5 'III' 



B. Track Relay 
T. Track Battery 
1 C. Clearing Relay 






^ iji 


, 3 









^r 4 


Z H- Electro- Magnet -for 


i * \ 


rj ' 



j 4 








D. for operating Distance falv 












i Battery 

^ I0 






P Battery 

^ 12. 








Return Wire 

Kig. 290. Diagram showing Circuit of 

The signals worked from mechanical boxes which are con- 
trolled by the "Track-Circuit" are provided with a controller 
and replacer similar to that in fig. 126. The signals stand 
normally at danger, and are lowered by approaching trains 
if the section ahead be clear. Each distant signal is " led " 
by the home signal it applies to, as well as being controlled 
by the upper arm (when fixed as a lower arm). In order 
then that a driver may find the distant signal " off " when 
he sights it, the stop signal say 14 in fig. 290 is lowered 
when the train is passing stop signal 10, providing of course 
that the line be clear. When stop-signal 14 is lowered dis- 
tant signal 14 follows, so that with sections of 1,200 yards a 
driver gets ample notice of the state of the distant signal. 
There is, therefore, no basis for the impression that with 
the " normal-danger " system drivers will continually be 
sighting signals "on," and be running expecting momen- 
tarily that the signal will drop, with the fear that some day a 
driver may be led into a trap. Nor will time be lost by the 
constant checking of trains. It must, therefore, be under- 
stood that if the section be clear drivers will generally find the 
signals " off " for them when they sight them. 

Fig. 290 is a wiring diagram of the Hall signals, as in- 
stalled on the North Eastern R. 

An electro-magnet or track relay B is connected by means 
of thick insulated copper wires 3, 4, to the entering end of 

Hall Automatic Signals, North Eastern R. 

overlap N of the section L N contact i on relay B 2 is open, 
which breaks the circuit of battery T 2 , and this in turn de- 
energises B 1 and holds signal 10 at danger. Directly train 
No. 3 passes off overlap N relay B 2 will be energised, 
closing contact i, and B 1 will restore its contact i, and signal 
10 will clear for No. 5 train. 

Assume that distant signal 12 has not for some reason 
been lowered, and that train No. 5 is approaching it. The 
distant signal cannot be lowered until after home signal 12 
is " off," therefore train No. 3 must pass the overlap of stop 
signal 14 (400 yards beyond the signal) to allow contact i 
on relay B to close. 

Contact 2 on relay B 2 having' closed when train passed 
off overlap N and train No. 5 holding contact i closed on 
relay C 1 , the circuit for stop signal 12 is completed thus : 
Erom battery 12, wire a, contact i, wire b, contact 2 on B 2 , 
wire c, electro-magnet H 12 , wire d, coils of relay P 2 , wire e, 
wire /, spring 4 (being closed because the stop signal arm 
10 is " off "), wire g, contact i on relay C 1 , wire h, electro- 
magnet D 12 , wire *', and " return " wire to opposite pole of 
battery 12. 

The circuit is now complete through both the home signal 
and distant signal electro-magnets, and also through the 
coils of relay P 2 . The resistance of relay P 2 reduces the 




Fig. 291. Hall Automatic Stop Signal with Lower Distant Arms, North Eastern R. Fig. 292. 



current from battery 12, so that there is not sufficient power 
to open the valve of either the home or distant signal. 
There is, however, sufficient power in relay P 2 to close con- 
tact i (as shown by dotted line), forming a path for the 
current to pass through H 12 to open the valve and admit gas 
into the cylinder and raise the signal arm. 

Springs 4, 5, are the switches t l , f 2 , fig. 290, which will 
close when the arm is off. The closing of 5 sets up a path 
for the current from H 12 to wire e, cutting relay P 1 out of 
the circuit, thus increasing the flow of current back through 
D 12 sufficiently to open the valve and lower the distant arm. 
For directly spring 5 closes by arm 12 moving to the " clear ' 
position the whole of the current is diverted from the coils 
of relay P 1 and the contact i opens, and the amount of cur- 
rent passing through H 12 also passes through D 12 . When 
the train passes signal 10 on to section L M it would short- 
circuit relay B and break contacts i and 2 and close 3. The 
breaking of i .will hold signal 8 at danger, the breaking of 2 
will restore signal 10 and distant 12 to their normal position 
"on," and the closing of contact 3 will hold stop signal 12 
" off " while train is running L N, and the closing of contact 
2 on B 1 will hold signal 12 "off" while train is running 

M N. 

All the signals on the North Eastern installation are of 
the Hall electro-gas pattern as illustrated by fig. 248. 

Figs. 291 to 294 are views of some of the signals. 

Figs. 291 and 292 illustrate stop signal 10 (fig. 289), with 
the lower distant arms for Bishophouse Junction. The upper 
arm is purely automatic, the left distant arm is semi-auto- 
matic, being automatic at all times, except when the signal- 
man wishes to keep it at danger. When he does, he puts 
back in his locking frame the lever set apart for this signal. 
The right distant signal arm is not automatic, except so 
far as being thrown to danger as a train passes it. Neither 
distant signal is coupled up in the usual way and worked 
by wire, but they are provided with gas tanks and cylinders 

like an automatic signal, and on the levers there are elec- 
trical contacts which complete circuits determining the gas 
valves on the signal. 

The post is the standard of the North Eastern R. for 
bracket signals, and upon it are fixed the pillars used for 
the automatic signals. At the foot of the signals are the 
relay boxes, which are fixed there and the batteries in a 
separate box, instead of having an upper and lower base for 
the dolls of bracket signals. On the left at the foot will be 
seen a crank which is coupled to the distant signal arms, 
and also to Raven's fog-signal apparatus. This apparatus 
consists of an obstruction in the "four-foot," which causes 
a whistle to be blown on the engine if the signal be at 
danger. The obstruction stands vertically when the signal 
is at danger, and the cylinder of the distant signal pulls it 
" off " when lowering the arm. 

Figs. 293 and 294 illustrate a standard two-arm auto- 
matic signal. The concrete base will be noticed and also the 
gas shute (immediately in front of the signal). The tub in 
the foreground is not part of the signal appliances. 

The small numbers on the signals are the numbers given 
to them for localisation purposes. The first figure is the 
mileage from York, and the second figure or letter the first 
or second signal in that mileage. On the down road figures 
(i. 2.) arc given, and on the up road letters (A. B.) These 
indications have since been made plainer by being painted 
I2ins. deep on the mechanism case. 

Electro-Pneumatic System. 

This system is also very commonly referred to as the 
VVestinghouse system and is controlled in Great Britain and 
the Colonies by McKenzie, Holland and Westinghouse, 
Ltd., and on the Continent by the Westinghouse Brake Co., 
Ltd., London. Electro-Pneumatic Automatic Signals are 
extensively used in this country, but almost entirely upon 
railways worked by electric power, and the system is there- 
fore described in the succeeding chapter. On the Lancashire 


Fig. 295. Circuit piagram, 






| *s^%fl 

^r Stop for B "fC 

DistentforC Sr>/&-C-ffl 

, " 4 ^--j = r ' i 

r* r^ i. L. 


m ~ Distant for A 


(g-5/(jD^r 5 ^ Distant for B 

Starting tbr B^ 

Storting fofC^ 


Fig. 296. Automatic Signals on the Paris, Lyons and Mediterranean Railway. 

and Yorkshire R., however, some Electro-Pneumatic Auto- 
matic Signals with " Track-Circuits " have been fixed on 
the main line between Middlcton Junction and Castleton. 
Automatic Signals on American Steam-Worked Railroad. 

The difficulty of dealing with this branch of the subject 
is its vastness. 

America is covered with automatic signals, and it is hard 
to single out any installation for especial mention. In Eng- 
land and on the Continent the installations are the reverse 
of numerous, and a few years ago the same might have 
been said of the United States, for as recently as 1899 the 
number of automatic signals was placed at 6,496, whereas 
now they are estimated to number more than 30,000. 

Special reference should be made to the Lehigh Valley 
RR., of which the main line, 440 miles long, was equipped 
with automatic signals some years ago. The Delaware, 
Lackawanna and \Yestern RR., the Chicago and Alton RR. 
and the Michigan Central RR. also all have notable instal- 
lations, particularly the last mentioned, which the Author 
believes has more double line equipped than any other railroad 
in the United States. 

On the Erie RR. 80 miles of double line on the New York 
division of the main line are provided with the block system, 
but Hall automatic signals, see fig. 295, are to be placed on 
this length. 

Automatic signals are being installed on a length of about 
500 miles of the Lake Shore and Michigan Southern RR. 
between Buffalo and Chicago. Reference was made in 
Chapter VII. to the Harriman lines. 

It must not, however, be assumed that when a line in 
America is equipped with automatic signals that it is inter- 
locked as on British railways. The contrary is, in fact, the 
case, for only at rare intervals are the points in the main line 
coupled to signal-boxes or even ground frames, and the 
remainder, even facing points, are worked by loose levers. 
Where automatic signals are provided the electrical connec- 
tions to them are threaded through the points, so that should 
the switches be interfered with the signals will go to danger. 
On the other hand, there is rarely any reciprocal interlocking, 
i.e., when the signals are '' off " they do not hold the 

Autojnatjc Signals, Erie RR. Fig. 295. 



There are about 53,000 route miles of railway protected 
by either the block system or automatic signals in America, 
so that there is still a large amount of line unprotected. 
Continental Installations. 

Progress in automatic signalling on the Continent has 
been slower than it has been in Great Britain. The reasons 
for this are set out in Chapter XL, and with the exception 
of a trial one in Austria there are only two installations on 
steam-worked railways, and these were laid down by the 
Cie. Electriqites Signaux pour Chemins dc Per, of Paris, upon 
the Hall system. 

C" ~) c^f^' C~ ~i f ^rd 

ii ii 

I.,..' i SEE i 




Fig. zq-j. Indicator Hoard for Automatic Signals, P., Lyons and M.R. 

One of these is on the P. Lyons and M. between Laroche 
and Auxerre St. Germain on the branch from Laroche to 
Cravant. The portion protected by automatic signals is 38 
kilos (2$6 miles) in length and the line is divided into four 
sections : Laroche to Bonnard ; Bonnard to Chemilly ; 
Chemilly to Monetau, and Monetau to Auxerre St. Germain. 

The line is double throughout and is equipped with 
' Track-Circuits," which average 1,200 metres (1,308 yards) 
in length. Each section is provided with a distant, stop and 
starting signal, of which an example is given in lig. 296. 

The distant signals are of the disc pattern as used in 
America, but the stop and starting signals are of the type 

ordinarily used in France, the American disc signals not 
having been adopted by the P. L. and M. Co. in consequence 
of it having been found that' the reflection of a rising or 
setting sun obscured the signal to a driver. Motors have 
therefore been attached to the existing signals. 

In diagram, fig. 296, there is a crossover road and a 
siding connection at B. The signals are normally at danger 
and are lowered stop, starting and distant as the train 
leaves the section in the rear. In fig. 296 a train from A 
has entered the A B section and one from C (in the opposite 
direction) is approaching the starting signal C. 

In order that the points shall not be worked when a train 
is in a section an indicator and continuous ringing bell are 
provided for each line as seen. These indicators show when 
a train is approaching, also when the section in advance is 
" clear." The bell rings when a train is approaching. There 
is therefore a visual as well as an audible indication given 
of the approach of trains. 

Fig. 297 icpresents the indicator board provided. The 
disc a when in view (as illustrated) indicates that the up (or 
down) line from the section in the rear is occupied and disc 
b shows that the up (or down) line for the section in advance 
is occupied. The discs c d act similarly for the rear and 
advance sections for the other line. 

Before the siding points on the up line may be used the 
chef-du-gare (station-master) has to take the key e 2 out of 
the box c, and this throws to, or keeps at, danger the sig- 
nals for that line. Key /* out of box f acts similarly for the 
other line. Box g with key jj- 2 acts for both lines, as when the 
crossover road is to be used and, as illustrated, the key has 
been placed in box /;'. 

The switches / I are for emergency use and will throw to, 
or keep at, danger the distant and home signals for either 
or both lines. Switches h m act on the same signals and on 
the starting signals too. 


On the main line of the Midi R. from Bordeaux to 
Toulouse, Hall automatic signals have been provided between 
the former city and Langon, a distance of 42 kilos (26 miles). 

The signals employed are of the Hall disc type. 




Timmis System. 

THE first installation of automatic signals in Great 
Britain and the first on a railway worked by electric power in 
the world was on the Overhead R. at Liverpool. 

The stations on this railway, being both numerous and 
close to each other, the estimate for the initial cost, and sub- 
sequent working and maintenance, for ordinary mechanical 
signalling, was so large that the engineers sought for some 
other method of meeting the Board of Trade requirements 
at a less cost, with the result that the automatic system in- 
vented by Mr. I. A. Timmis, of Westminster, was adopted. 

Fig. 301. Timmis Long Pull Magnet. 

The line consisted of the two terminal stations, two junc- 
tions, and 15 intermediate stations; and of these, 13 stations 
are protected by automatic signals, there being two signals 
for each line a home signal about looft. in the rear of the 
station, and a starting signal in advance, the starting signal 
acting as the " distant " for the next station. 

The line is not on the " Track-Circuit " system, but is 
controlled by electrical contacts, and the signals are on the 
" normal clear " method. Each signal is put to danger as 
the train passes it, and when a train enters a station it puts 
the home signal at the station to danger, and lowers the 
home signal at the station in the rear. When it leaves a 

station it puts the starting signal there to danger and lowers 
the starting signal at the station in the rear. Each train is 
therefore protected by two absolute stop signals. 

The locking and unlocking device is not of the usual 
electrical treadle arrangement, but is a lever by the side of 
the line, and contact is made by the lever being struck by 
a bar attached to the last vehicle on the train, so that the 
section is not cleared should the train have broken loose, 
and the whole of it not have passed out of the section. 

The signal posts are iron and of the ordinary pattern, the 
upright signal-rods being connected to one of Timmis' long 
pull magnets, which are fixed in a case on the front of the 
signal post. 

The principle of the long pull magnet may be seen on 

Fig. 302. Making Contact for Timmis Signals, 
reference to fig. 301. Fixed on the base a is a magnet b b, 
in the centre of which is the core c of an armature d, which 
is attached at e to the upright rod connected to the signal. 
The magnet is also surrounded by a case / / forming part 
of the armature. 

On the magnets being energised by a current being sent 
through them, set up by a train striking a " making con- 
tact," they attract the armature by means of the core and 
case, and the upright rod is lowered, and as the armature 
gradually approaches the magnet, it gets into the field of 
influence stronger and stronger, and by this means the load 
is moved at a much smaller expenditure of energy. Each 




Fig. 303. Breaking Contact for Timmis Signals. 

signal is lowered by a current of 5 amperes at 40 volts pres- 
sure, and when it is "off" a very large resistance is switched 
into circuit to reduce the current to about '25 of an ampere, 
which is sufficient to hold the signal "off." 

The " making " and " breaking " contacts are separate 
devices. The " making " contact is illustrated by fig. 302. 
It is fixed to the side of the line, and consists of a lever, a, 
connected to a commutator b, and on a train passing the con- 
tact, the lever a is turned to the right by the striking plate, 

f Sirtluiq Bar 
(alfachtX to Train, 

^ Contact 

(at litLe of ' ^ 

Direction ofTr 

Fig. 304. Striking Bar for Timmis Signals. 

carried by the last vehicle of the train. When the lever is 
so turned the commutator is moved, so that the spring c is 
put in contact with the insulated piece d, and a current flows 
through the magnet c of the " making " contact, through a 
mercurial contact on the signal immediately in the rear (so 
guaranteeing that it is at danger) to the long pull magnet on 
the signal that may now be pulled " off." 

In fig. 303 is seen the " breaking " contact, which has a 
similar lever a connected to an armature b, but the spring 
c is normally in contact \vith the insulated piece d, and when 

'Jp line Wire 

the striking plate on the last vehicle causes the commutator 
to be turned, the current is reversed and the signal being 
released, it goes automatically to danger. 

Fig. 304 shows the striking plate, and its action is clear. 

Fig. 305 is a view of the signal case, and shows the 
switch, whereby the current is reduced from 5 amperes to 
'25 ampere, sufficient to hold the signal " off." 

The switch a is inserted in the upright-rod b of the sig- 
nal, and when raised by the action of the magnet, situate in 

Signal Case. 

the lower part of the case, the armature r is attracted and 
held by the magnet d, until the current is reversed by the 
train going over the contact "breaker" when the switch 
is released, and the signal rod falls by its own weight. 

Fig. 306 is a diagram of the electric wiring, and it in- 
dicates the positions of the "making" and "breaking" 

At the Station A, there is on the up platform a lever for 

UD tine rvt 

, ^ ( 


'v! * 



Platform ^Batterj J 'V "* I "*" rry-e 

_^>^Z- ; i '.' i i * "* : -"-'ii i 


rj ~^ 



1 1 

btneral Negative ^ _^ 

( i.^. -i -^ ~ -i j> * ' 


6.^,01 Xsot,, 


\ * 

a Healing Contact 
t Making Contacl 
l Magnet tio*e$ 


_.* j _ __Jo Breaking Contac 


Fig. 306. Wiring Diagram for Timmis Signals. 



working the cross-over road, through the frame a of which 
it will be observed that the electrical wires are threaded. 
This is in connection with the complete arrangements made 
for the protection of the line when the cross-over has to be 

Before the lever can be moved a key has to be taken out 
of each home signal the up and the down and inserted in 
the frame. These keys cannot be withdrawn if a train be in 
the section, or a signal be " off," and the withdrawal breaks 
down the circuit and prevents any train approaching, and 
not until the cross-over road has been restored to its normal 
position, and the keys withdrawn from the frame and re- 
placed in the signals, is the section free. 

Panton's Illuminated Semaphore Arm. 

Reference might here be made to a new form of arm now 
being tried on the Overhead Railway. 

With the object of rendering signals more distinct, parti- 
cularly at night, Mr. J. A. Panton, of the Liverpool Over- 
head R., has recently patented an arrangement by which the 
semaphore arms are illuminated by a line of electric light 
which may show any desired colour. The signals are in- 
tended for electrically operated railways, though they mav 
obviously be used wherever a supply of electricity is avail- 

It will therefore be observed that the line of light on the 
arm is provided in addition to the present coloured spec- 
tacles and also that the reversal of the face of the arm 
causes the green (or white) side to be presented to the driver 
when the signal is "off." The arrangement would also 
appear to be a very suitable one for indicating conspicuously 
the difference between " home " and " distant " signals. It 
will also be noticed that extreme simplicity is a distinctive 
feature of the invention, of which the principal advantages 
claimed may be briefly summarised as follows : (i) The 
greatly increased conspicuousness of the visual signal ; (2) 
the change in the position of the coloured lights in addition 
to the change of colours; (3) in daylight the reversing arm 
presents its white or green face when lowered ; (4) the warmth 
of the lamps and the reversing prevents the signal from 
becoming obscured by snow. 

Tim m is-Lavc ra zzi Signa Is . 

The railway in the grounds of the Paris Exhibition of 
1900 was fitted with signals of the Timmis-Laverazzi 

The " breaking " contacts were identical with that illus- 
trated by fig. 303, but the " making " contact was modified 
as shown bv fig. -507. 




: :fi 


\'\g. $uba. Panlon's Illuminated Semaphore Arm. 
On reference to fig. 306*1 it will be seen that the spectacle 
casting is in two portions, one portion carrying the spec- 
tacles, whilst the other portion carries the wooden arm, and 
can revolve upon a trunnion,, which is mounted in a bearing 
in the spectacle casting. Fixed on this trunnion is a pinion 
wheel which gears with a rack fixed upon the post, conse- 
quently as the arm is raised or lowered this pinion causes 
the outer wooden portion of the arm to make half a revo- 
lution. Two separate and insulated contact pieces connected 
to the lamps on the arm engage with two electrically con- 
nected contacts on the post, whereby the lamps are lighted, 
only one colour being lighted at a time, and that only when 
the arm is in its correct raised or lowered position. 

307. Making Contact, Timmis-Laverazzi System. 
On the striking plate (also used as on the Overhead line) 
coming in contact with the lever a. the latter is turned to the 
right, and the roller b is pressed against the spring c, raising 
the armature d against the magnet e, \\ hich, becoming ener- 
gised, holds up the armature and so keeps the circuit closed 
until the current has done its work, and lowered the signal 
to the " off " position. This ensures that the current is on 
sufficiently long to get the signal " off," as the contrary 
might happen if reliance had to be placed only upon 
the switches being placed momentarily in contact by 
the action of the lever a. 

The signals are not of the usual semaphore pattern, but 
are of the " banner " type shown in fig. 308. 

Carried on a wrought-iron post is a circular frame a 
with a sheet of plate glass b in front, and opal glass c at the 
back. Inside the frame, and balanced in the centre on a 

M 2 



Fig. 308. Timrnis-Laverazzi Signal. 

Fig. 3090. Sykes' Electric Tunnel-Signal. 

dering the disc clearly observable, whilst the opal glass 
makes a good background for the day view. 

Fig. 309 is the wiring diagram for the line and general 
details of the electrical connections on the signal, and the 

pivot, is the signal disc d, which is made of fine red calico, " breaking contacts " and " making contacts, 
carried in a light frame. 

The electrical fittings are in the base of the outer frame a, 
and behind the cross pillar c is a small long-pull magnet 

Waterloo and City. 
This is an electrically operated Tube railway, and whilst 

connected to an upright rod, the other end of which is at- it is neither protected by automatic signals nor " Track-Cir- 

tached to the signal disc. On a train passing the " making cuits," yet the signalling is sufficiently interesting and of 

contact," a current is sent to the magnet, which, being mag- such a nature as to demand a place in this chapter, 

netised, raises the upright rod, and the signal is placed in j ne mechanical work was done by the Railway Signal 

the " clear " position as shown in dotted lines in fig. 308. c o _ and the electrical by Mr. Sykes, whose train protection 

In the rear of the frame there is a lamp, which, shining and fouling bars, detectors, selectors and " lock-and-block " 

on the opal glass at the back, gives a good clear light, ren- are employed throughout. 

line nire to Signalin rear Line rvire_ to Jhgnal ^1_ _ Line ni re to Signal 6 

j[tJTfe resistance is placedin the Pignut Case 

I\The resistance is placed in the Signal Case 

,;J_XJ * 




General Negative 




To front contact 
fo back can tact 


line wire 1o Signal 
in rear ir~ 

Ib frvnt contact 
To bacK contact 

tine itift from Mjkinq fontuft , r ~ . c j 

L - Fig. 309. Wiring Diagram I immis-Laverazzi hystem. 



The line is ii miles in length, double throughout, and the 
only stations are at the termini. 

About the middle of the tunnel there is on each line an 
advance signal with two lower distant discs. These are aJl 
operated electrically, the current to the advance signals being 
set up bv a lever in the locking frame, but the distant signals 
are actuated by a current set up by the lowering of the cor- 
responding stop signal and the advance signal above the dis- 

Fig. 3090 shows the signal employed. The incandescent 
lamps B C are in duplicate, and between them and the bull's- 
eye is a screen with a red and a green glass. The screen is 
attached to a lever S on which there is a revolving armature 
H working between the coils A A. Normally the lever rests 
on contact E, and through K current enters by terminals O P 
to the coils A, energising them and turning the armature H 
so that the lever S is raised as far as the stop J allows. The 
lever S then makes contact with D, which causes Off to be 
recorded in the signal box (E will record On). \Yhen " off " 
the contact L joins up F G, which in the case of the advance 
signal sends current to the lower distant signals. At the end 
of the lever S is a weight K, also a screen X, which, when 
" off," blocks out the red light M. 

Automatic train control is provided for all the signals on 
the \\~aterloo and City, whereby the current is cut off from 
the train in case a signal be passed at danger. 

Boston Elevated Railroad. 

The provision of " Track-Circuits " on electrically worked 
railways is an altogether different problem to their adoption 

on steam-worked railways owing to the running rails being 
sometimes required to return the power current and to the 
probability of the power current interfering with the signal- 
ling circuits. 

The first railway that had to deal with this problem 
was the Elevated RR. of Boston, Mass., U.S.A., of which 
Mr. H. G. Brown was the signal engineer. He devised a 
system which surmounted the difficulty, and which has, with 
improvements, been adopted by the \Vestinghouse Brake 
Co., Ltd., and installed by that company on the Metropolitan- 
District and other railways. Mr. Brown's patent is described 
subsequently (p. 174), and illustrated by fig. 324. 

The Boston Elevated RR. was opened in 1901, and was 
signalled by the Union Switch and Signal Co. Except at 
the junctions it is signalled automatically. 

The trains are worked by electric power on the " third- 
rail " system, and the signalling is on the same principle as 
on an ordinary railway, except that the running signals are 
of the dwarf pattern. 

There are seven signal boxes on the line. A, B, C, D, E, 
F, G, all of which have electro-pneumatic power plants 
similar to those described in Chapter XIX, except boxes 
B and E. 

The connections worked from the signal boxes are : 

A. 25 signals, 2 two-arm signals, 28 switches. 

B. 7 signals, 3 facing point locks, 3 switches. 
C- 12 signals, 5 two-arm signals, 12 switches. 
D- 3 signals, 2 two-arm signals, 6 switches. 

E. 7 signals, 2 facing point locks, 3 switches. 

F. 7 signals, I two-arm signal, 5 switches. 
G- 12 signals, 2 two-arm signals, 10 switches. 

Fig. 310. 

Electric-Pneumatic Automatic Block Signal 
with Automatic Train Stop Attachment, Boston Elevated R. 



Fig. 311. 
Electro-Pneumatic Automatic Block Signal in Tunnel with Automatic Train Stop Attachment. 

Four compressors arc provided for 
motive power for the automatic signals 
and for the five interlocking instal- 
lations. The signals are fixed so as to 
allow for a one minute train service at 
an assumed speed of 30 miles per hour, 
the exact positions of the signals being 
determined by the situations of the 
stations and the curve of the line. 
Each car is provided with an auto- 
matic stop for putting on the con- 
tinuous brake in case a driver runs by 
a signal at danger. This arrangement 
is illustrated by fig. 310. 

After the explanations given in 
Chapter XII. the working of the signal 
will be understood, so that all that is 
now necessary is to explain the stop 

Suspended from the car is a valve 
a connected to the continuous air 
brake, and outside the near running 
rail is an arm b fixed on a shaft c i^ins. 
square, the other end of which is 
coupled to a piston in the signal 
cylinder d. \\ hen the signal is at 
danger, as in fig. 310, the stop is up; 
but when the signal is lowered, the 
rod e, coupling the shaft c to the 
signal, is depressed and the stop arm 
drawn clear of the air brake valve. 
Should a driver then overrun a signal, 
the stop rirm, being up, would open 
the air brake valve and pull up the 

The signal is fixed on the staging 
at the side of the line; but in the 
underground portion of the line there 
is not clearance and consequently 
other arrangements had to be made, 

ftaf* 't!t> r-ff - 


! 3 j 5 * 

Fig. 312. Diagram of Operation of Electro-Pneumatic Block Signal System, Boston Elevated Railway. 



Fig. 3.3- Circuits controlling: Locking Devices to Charles 

River Draw Bridge. 

B - 



ISL. 2OK. Bridge signals. 




t.2.3.4: Circuit closers attacttstf to gates 

and these are illustrated by fig. 311. 

It will be noticed that the third rail is in the " six foot," 
whilst in the open (fig. 310), it is outside the running rails. 
Each car carries two slippers .v 1 , x", one on each side, so that 
current is obtained at any point on the road. A recess is 
made in the wall of the tunnel to receive the signal which has 
its arm removed, as it is not necessary, as the signal arm 
cannot be seen in the tunnel. The part marked a is the por- 
tion of the spectacle to which the signal arm is connected. 
As the base of the signal is considerably above rail level, the 
automatic stop arm cannot be actuated by a shaft as in 
fig. 3'io, and, therefore, the arm is moved by an independent 
cylinder b fixed in the "four foot." This cylinder is con- 
nected by |in. pipe to the signal cylinder, and when the 
signal is lowered air is admitted to the stop cylinder which 
draws the stop arm clear. On the signal returning to dan- 
ger, air is admitted to the other end of the step cylinder 
and the arm is raised again. 

The arrangement of line wires is shown in fig. 312. The 
signals are worked on the " normal clear " method, whereby 
the signal immediately in the rear of a train is put to danger, 
and the signal admitting a train to the section in the rear 
is lowered providing the line be clear. Trains therefore 
find the signals " off " as they approach them, unless a train 
be in the section ahead or a rail be broken. In fig. 312 the 
section extends from one. insulated joint a to another b, and 
the train being in the position shown in the illustration has 
passed over the second insulated joint ft, and thus put to 
danger the signal B immediately behind it, and lowered the 
other signal A. The insulated joints at Boston are fixed 
180 ft. in advance of the signal, so that there is an " over- 
lap " of that distance. 

The Boston Elevated R. passes out of the City proper 
over the Charles River, and to allow for .the passage of 

conductor cable H Satin 'C' 
(Lever - II., 

Circuits on Bridge and in Engine Room 

controlling Lock to Dra-M Bridge 

operating Devices. 

shipping, a draw-bridge has been provided. This is con- 
tiguous to signal box C, and fig. 313 shows the electrical 
connections by which the draw-bridge is controlled from C- 
No. 21 lever in that cabin is the draw-bridge control lever, 
and before that lever can be moved the signals worked by 
levers i8L and 2oR have to be at danger. When No. 21 
lever is moved, air is admitted to cylinder P in a similar 
manner to the point mechanism described in Chapter XIX, 
which withdraws the bridge lock plunger, and raises the lock 
L on the valve of the draw mechanism in the draw-bridge 
tower. The man at the bridge then withdraws the draw- 
wedges. \Yhen the work is completed and the bridgeman 
has restored the bridge, he must put back the draw wedge 
or the return current to lever No. 21 will be broken at G, 
also the bridge must be in perfect alignment at F or the 
current will be broken there, and consequently the signal- 
man in C will not be able to complete the stroke of lever No. 
21, and therefore he will not only know that the work is not 
properly completed, but signals i8L and 2oR will be inter- 
locked at danger until lever No. 21 is properly back. 
Actuated by the lock plunger P is the circuit closer B, and 
as the bridge can be turned right round and not necessarily 
reversed, there is a corresponding circuit closer at A, and an 
alignment indicator at D to correspond with E. 

North Shore of San Francisco. 

1 his is an old narrow gauge steam railway on which 
standard gauge electric trains run, an additional running 
rail having been provided, and the line has an extra interest 
in the fact that power for working the electric trains and 
for signalling purposes is obtained 180 miles away in the 
Sierras, where, at the Colgate power house, there is a water 
head several times that at Niagara. 

The signals arc worked by storage batteries charged 
from the live rail and not by primary batteries. 



Both steam and electric trains run over the railway and 
the signals protect both. At night when the electric supply 
is shut down the signals are actuated by the storage 

There arc 58 electric motor signals on this length of 10 
miles and 32 sections of "Track-Circuit." The installation 
was carried out by the Union Switch and Signal Co., who 
tried here, and subsequently adopted, Strublc's system. 

Si ruble's System. 

The principle of this system lies in the use of alternating 
current for the "Track-Circuits," as it is able to induce a 
current in another circuit brought within its magnetic field. 
The track relay is, therefore, of the induction type and re- 
sponds to alternating current and not to direct current. An 
excess of direct current cannot cause a wrong operation of 
the signal, other than to cause it to go to danger, for if a 
fuse or other protective device fails to open the circuit, the 
relay coils will be destroyed. With this relay there is no such 
thing as residual magnetism, and the points of pick-up and 
of release are identical, except as affected by twice the 
mechanical friction of the moving parts. 

At a meeting of the Railway Signal Association of 
America, held in New York in May, 1904, the system was 
described as follows : 

Two main feed wires bearing alternating current at, say, 60 cycles 
and 2,000 volts, extend the length of the system, and across these are 
connected the primaries of the track transformers, the secondary leads 
of which are connected across the rails at the exit end of each track 
circuit. Across the rails at the entering end are connected the terminals 
of the induction relay. \\'e now have a circuit consisting of the secon- 
dary of the transformer, the rails and the coils of the track relay. 
Through this circuit passes simultaneously two kinds of current, alter- 
nating, induced by the primary of the transformer, and direct, the return 
from the car motors. 

Since the direct-current tends to make ineffective the alternating 
current, an impedance coil is connected across the relay terminals, or 
track rails ; this has low ohmic resistance, but high inductive resistance 
or impedance to the passage of alternating current, and serves the 
purpose of shunting the direct-current from the relay, while compelling 
the alternating-current to pass through the relay. 

New York Interborough R. (the Subway). 
One of the more remarkable successes connected with the 
construction and operation of the New York Subway is the 

signalling. To Mr. George Gibbs, the consulting engineer 
responsible for the rolling stock and signalling, is due the 
credit for the main ideas of the scheme, and the Author has 
to acknowledge that he never met with work demanding 
more care in design, where new departures were so abso- 
lutely necessary and where the result has been so complete 
in practicability coupled with simplicity. 

Mr. Gibbs was assisted by Mr. J. M. Waldron, the 
Interborough Signal Engineer, and by the contractors, the 
Union Switch and Signal Co. 

There was no similar work anywhere from which experi- 
ence could be sought. What was suitable for the elevated 
roads in New York and Boston would not do for the Sub- 
way, nor could anything be learnt from the signal arrange- 
ments of the underground lines in London and Paris, as 
there were no express lines there, and the rate of speed was 
much lower. 

Practically everything had to be especially designed, and 
when the scheme was adopted it was found that existing 
signal and point constructions would not be suitable owing 
to want of space, and they all had to be re-designed, too, 
except the locking (detector) bar and fastenings. 

The general scheme of signalling is that the express 
tracks are signalled throughout, and the local lines at those 
places where there is an imperfect view of the line. The 
signals are automatic, with "Track-Circuit," on the West- 
inghouse electro-pneumatic system, by which the switches 
are also worked. 

Distant signals are provided, and an "overlap" is allowed, 
to which reference will be made subsequently. 

Alternating current "Track-Circuit " is employed. This 
system is the one just described as in use on the North Shore 
of San Francisco steam-electric railway, and its adoption on 
to the Interborough R. has proved to be quite satisfactory. 

Another feature is an automatic stop by which the train 
is pulled up in case a stop-signal is passed at danger. These 
were made by the Kinsman Block Signal Co. (see Chapter 

Bowltnq Green 

nail S1ree^ 
I..-., * 


i HMh St. Cana/St. 
II fc j 

{'"?,J B/eekcrSf. ^ 
, UJL^ >^ -- Local ^^ 

^JL -^ ^ 

, ^ ED^/lL^ ^^ ^ impress ** i ^>'i -^fct_ 
-r ^^^L, 

-rr ~*r 

W H 1 1 


1 ^ww-a ^ 3 

^ P 3 





| ~] 


Fig. 314. Diagram of Automatic Signals on the New York Subway. 

A I - T OMATI C S 1 G .V ,1 L S O .V ELECTRIC R A 1 L TIM 5" .S . 


Fig- 3 '5' Front View of Slop and Distant Signal. 

XVII.), and were found particularly useful during the strike 
of the Interborough employees in the spring of 1905. An 
interesting feature of the stop is that after it has been low- 
ered clear by the stop-signal being pulled " off " it does 
not rise again when the signal is put to danger, but is held 
off by an electric current until the whole of the train has 

The signals are operated on the " Normal-Clear " prin- 
ciple, and the stop-signals have red and green lights and 
the distant signals show yellow when the signal is " on " 
and green when at clear. 

The lengths of the "overlaps" have been arranged on 
scientific principles, and regard has not only been had to the 
maximum speed and braking conditions but also to gradients 
and curves. Tests were made to arrive at these data, and 
the length of each "overlap" was determined by them. The 
length of the "overlap" in turn governed the length of the 
block section, which was twice the length of an "overlap." 
If an overlap had to be Sooft. the block section was i,f>ooh. 
In each section there are two stop-signals, the length of the 
"overlap" being between the stop-signal in the middle of the 
section and the stop-signal at the end of the section. As a 
block section would extend from stop-signal A to stop-signal 
C, with signal B midway, it follows that a preceding train 
must have passed C before a following train could leave A. 
Signal B would be at danger if the first train had not passed 
signal D, and therefore should the second train overrun 
signal B the train would be automatically pulled up, and as the 
length of the "overlap" had been determined by the gradient 
and curve and the braking distance increased by 50 per 
cent, as a margin of safety- for a train at 35 miles per hour, 

it would have plenty of room to come to a stand before 
reaching signal C. This novel feature, which was intro- 
duced by Mr. Gibbs, should recommend itself to other 

Between certain stations there are third lines of rails, 
which are used for south-bound (down town) traffic in a 
morning and north-bound in the evening. When used for 
south-bound traffic all connections giving access north-bound 
are electrically locked at the signal-boxes at each end so as 
to prevent opposing movements. This control is reversed 
in the evening. 

The signal work was carried out in a remarkably short 

Fig. o ll >- 

View of Stop and Distant Signal. 



Fig. 317. Stop and Distant Signal, New York Subway. 

space of time. The work was not com- 
menced until June, 1903, and yet all was 
completed and ready for work when the line 
was opened on October 27th, 1904. That 
the contract was ably carried out is proved 
by the fact that during the month of Feb- 
ruary, 1905, there were 4,206,720 signal and 
automatic stop movements and only 13 
failures, or an average of one failure for 
323,594 movements. The average number 
of movements has increased to a monthly 
average of 5,282,028, but the rarity of failure 
is still prominent. 

These good results could not, however, 
have been obtained without excellent over- 
sight and constant attention by the mainten- 
ance gang. The railway company attribute 
this success to the nightly inspections which 
are made of the points, signals, automatic 
stops, junction boxes, relays, impedance 
coils, transformers, lamps, fuses and track 
wiring. The insulated joints are also swept 
every night to remove all conducting 
material such as splinters from brake shoes, 

Fig. 314 is a diagram of the line from the 
South Ferry Station to 28th Street. 

Signals worked from towers (signal-boxes) 
are shown in the "on" position, and auto- 
matic signals are shown in the "off" posi- 

The four lines of way commence at Fulton 
Street. At Brooklyn Bridge the City Hall 
loop has its connections. This line runs 
under the main tracks, and, as will be seen 
from the diagram, the facilities for passing 
from one track to another are ample. 

As already stated, the signals in the Sub- 
way have red and green lights for the stop- 
signals, and vellow and green for the distant- 
signals. The distant-signals arc always 
fixed under stop-signals and controlled 

Fig. ^15 is a photographic view of a stop 
and distant-signal. They indicate their state 
bv coloured lights, but miniature arm indi- 
cators arc provided in case the lights fail. 
As the tunnel is brilliantly lighted through- 
out these small arms are effective. The lower 
portion of the signal contains the actuating 
mechanism. The upper portion is divided 
into two parts. The rear and top 
portion contain the lense and indicator 
for the stop-signal. Then the distant- 



signal lense and indicator is placed in the other portion which 
is in front. The lenses are white and the colours are given 
by glasses in a frame inside the case, which are raised and 
lowered by rods connected to pistons in pneumatic cylinders. 
Each signal has two 4 c.p. incandescent lamps, the second 
lamp being provided in case one burns out. 

Ry the side of the right rail will be seen a trip. This 
is the automatic stop, which is actuated by a pneumatic 
cylinder fixed under the cover seen between the rails. When 
the stop-signal is lowered air is admitted to the cylinder and 

the trip is lowered. Should the stop-signal be passed at 
danger the trip comes in contact with a trigger 'on each car 
and thereby applies the continuous brake. 

In fig. 316 is a photographic view of the rear of a signal. 
It consists of four main parts. At the top is the transformer 
by which the high voltage alternating current at 500 volts 
potential is transformed down by a special double-secondary 
oil transformer. The second contains two coils, one of which 
supplies current for the incandescent lamps (already referred 
to) in the signal, and the other delivers current at 10 volts, 

Section f. G. 

Fig. 318. 

Dwarf Signal, 

New York Subwav. 

Section O.E.. 

which passes through the non-inductive resistance of i ohm, 
seen in the upper part of the case, to the alternating current 
n-lay in the lower part of the case. At the bottom is the 
controlling mechanism for the automatic train-stop. This 
mechanism is provided with a release, so that in case the con- 
trolling apparatus fails, the trip may be lowered clear to pre- 
vent unnecessary application of the brakes. This release is 
opened by a key carried by each trainman, who must hold 
the key turned in order to effect the release. As soon as 
he lets go of the key the trip will rise again. 

Fig. 317 is a drawing of the signal shown in fig. 315. 



- 3'9- Point Mechanism, New York Subway. 

The upper lense a is for the stop and the lower lense b 
is for the distant signal. The upper miniature arm a 2 is for 
the stop arm and the lower, b 2 , for the distant. The colour 
indications, red and green, for the upper signal are carried 
in the frame c and those for the distant in frame d. These 
are respectively raised and lowered by the rods c- d 2 which 
are coupled to pistons in cylinders c 3 and d 3 . Air is admitted 
from the main e to these cylinders by means of the usual 
Westinghouse valve c 4 d 4 controlled by electro-magnets c 5 
d 5 , in the case of a stop signal by the "Track-Circuit," and 
in the case of a distant signal by its stop signal in advance 
and by the upper arm. The miniature arms are operated 
by cranks c 6 d 6 which have a pin running in a slot in the 
frame. The drawing shows the slot d 7 for the miniature arm 
for the distant. Attached to the frames are cranks c 8 d s 
coupled to a slot in which are circuit breakers which oscil- 
late as the slot is raised and lowered and so make and break 
contact. Only those, c 9 c 10 , for the stop signal are shown, and 
these lead to the operation of the automatic stop, the release 
of the lower distant arm, the corresponding distant in the 
rear, and to cut off the air supply. 

Fig. 318 is a sectional drawing of the Interborough 
dwarf signal, which is actuated in the same way as the other 
signals (fig. 317). 

Space did not allow for the usual electro-pneumatic switch 
and lock movement, and a special design had to be devised. 
This is illustrated by iig. 319. The working of the 
mechanism by means of the pneumatic cylinder is fully 
described in Chapter XIX., fig. 387. 

The first movement takes a lock out of the rods b b at- 
tached to each switch, and by means of the slot c causes the 
rod d to move sufficiently to raise the locking (detector) bar 
to guarantee that no vehicle is on the switches. By the time 
this has been done the pin e has travelled along the straight 
portion of the slot /, and its continued travel causes the pin 
e to carry with it the rod g and consequently the switches. 
When that operation is completed the slot c gives a reverse 
movement to the rod d so that the locking bar again falls. 
At the same time the lock enters the cross rods b b, and the 
switches are again locked. 

It may be of interest to here state that during the first 
year the Subway was opened, and which ended on October 
27th, 1905, the number of passengers using the subway 
averaged 300,000 per day. 

The Author is indebted to the Union Switch and Signal 
Co. for the drawings and photographs published herewith, 
and to Mr. J. M. Waldron, the signal engineer of the road, 
for the diagram of the line and statistics as to operation, etc., 
quoted above and elsewhere in this book. 

Metropolitan-District Railway. 

This line, which is electrically worked, is equipped 
throughout with automatic signals controlled by " Track- 
Circuits " and Brown's relays, whilst the points, and signals 
at signal boxes, are operated by power, upon the Westing- 
house electro-pneumatic system. There were formerly be- 
tween South Kensington and Minories Junction (but ex- 
clusive of those stations) 13 signal boxes, and now there is 
only one at the Mansion House, besides an emergency one 



at St. James's Park, and whilst throughout the railway the 
majority of the signal boxes have been abolished, 29 still 
remain ; but of these, 5 are emergency boxes, and are 
opened only occasionally to work crossovers or siding con- 
nections, should irregularities in the traffic demand this : 13 
ol the remaining 24 have been equipped with the Westing- 
house electro-pneumatic apparatus for operating points and 
signals by compressed air. 

In those cases where the old locking frames remain, and 
the points continue to be worked by manual power, the sig- 
nals are not coupled to the usual signal wires but are actuated 
by power, the controlling current for which is switched on 
by means of contacts on the signal levers in the locking 
frame. There are, consequently, no wires to adjust, and the 
signals come " off " to the correct angle and go to danger 
properly. All the signals on the line are therefore actuated 
by compressed air. Where necessary, running signals, i.e., 
those governing the running of trains on the main line, have 
their respective levers in the locking frame controlled by 
the " Track-Circuit " for the line or lines they protect, so 
that whilst the signalman can put the lever sufficiently far 
back to replace the signal to the danger position, he cannot 
put it fully back until the train has passed the fouling point. 
The signal lever, not being fully back, " holds the road," 
and prevents the signalman from inadvertently pulling over 
the levers of conflicting points and signals. At interlockings 
the signals are automatically restored to danger by the train, 
independently of the action of the signalman, but he must 
replace the signal lever in the locking frame to normal before 
the signal can be again lowered to safety. Fouling points 
on sidings and on diverging and converging lines are pro- 
tected by the " Track -Circuit," so as to guard against any 
train or vehicle standing foul. 

The average length of the block sections through the 
tunnels (eastward of South Kensington) is gooft. On other 
parts of the line they vary from cjooft. to 4,oooft. Each 
signal is governed by the section immediately in advance of 
it, and this section commences 4Ooft. beyond the signal, and 
extends to 4ooft. beyond the next stop signal. This distance 
of 4Ooft. is the "overlap," and a signal cannot be lowered 
until the whole of the preceding train has gone out of the 
section in advance, and has also passed the next signal by 

There are no distant signals, except where motormen 
get a bad view of a stop signal. Where distant signals are 
provided they have yellow glass for the "on" indication 
and green for "off." All signals, except those actuated 
from signal cabins, stand normally "off." With such a fre- 
quent service, had the " normal danger " position been 
adopted, the signals would necessarily have had to stand 
longer in the clear than in the danger position. 

In the open, semaphore signals of the ordinarv pattern, 
fig. 320, with corrugated steel arms, are used. The spindle 
for the arm, the lamp bracket, and the motor, are all carried 
by one casting which can be secured to a post or wall by 
four bolts a good example of compactness. Where 
semaphore distants are provided the front of the arm is 
painted yellow. 

In the tunnels dwarf signals, as illustrated by fig. 321, 

Fig. 320. \Yestinghouse Electro-Pneumatic Signal, 
are used. When at danger, the upper spectacle (red) of the 
screen is before the lamp ; but when the signal is cleared air 
is admitted to the cylinder inside the box casting, and the 
screen is raised so that the green spectacle appears before the 
lamp. In order to place the signal at danger, the air is re- 
leased and the screen falls owing to its weight. Fig. 322 
is a view of a tunnel signal in situ: the automatic stop is 
shown, and also, in the rear, the relay and resistance boxes. 

The signals are lighted by gas, but each is also supplied 
with an Adlake oil lamp, which will burn continuously for 
a week without attention, in case the gas should fail. 

The train stops, fig. 323, are similar to those (already- 
described) in use on the Boston Elevated Railway (fig. 310). 

There are motor generators (in duplicate) at each of the 
eleven sub-stations to supply the power for the " Track- 
Circuits," charging the accumulators at the power-worked 
interlockings, and working r.ll relays, special circuits, and 
all other purposes connected with the signalling. The 
negative main of the signal system is laid the whole length 
of the railway and <So volts is maintained between it and the 
positive running rail. 



Fig. 321. Electro-Pneumatic Tunnel Signal. 

The power required for the " Track-Circuits " is 0*3 kw. 
per mile of single line. 

The diagram, fig. 324, shows arrangement and connec- 
tions of the "Track-Circuit" system. One of the rails is 
electrically continuous through the complete length of the 
installation, and is the positive conductor, the other rail is 
divided into block-sertions by means of the usual insulated 
rail-joints, all of them being connected to the negative 
signal-main. Resistances, R R 1 are inserted in these con- 
nections to reduce the potential difference between the rails 
to between 2 and 4 volts to suit the length of the block or 
other local conditions. 

\Yhen the block is unoccupied the positive current flows 
from the generator, through the continuous rail, the two 

relays, A and B, one at each end of the block, and the bal- 
last, all in parallel to the sectional rail and thence through 
the resistance, R, to the negative main back to the generator. 
When a train enters a block the current passes with prac- 
tically no resistance through the wheels and axles from one 
rail to the other, the relays are shunted and the signal 
allowed to go to danger. The resistances R prevent the 
generator being short circuited when the " Track-Circuit " 

Kig. 323. Electro-Pneumatic Train Stop, District Railway, 
is shunted, and are so proportioned to the combined resist- 
ance of the road bed and two relays in parallel that the 
shunting of road bed resistance cuts out but a small per- 
centage of the total circuit, so that the current increase in a 
circuit when shunted is not great, and this is important, as 
the track potential should be kept as nearly as possible con- 
stant. Weather conditions directly affect the road-bed 
potential : broken stone is the best ballast from an electrical 
point of view and cinder (slag) the worst. But it is worth 
noting that the flooding of the line has not been found to 
interrupt the working of the signals. 

The diagram, fig. 324, also shows the circuits of the 
relays R R 1 which control the signal circuits. The track 
co ; ls of the relays are permanently connected across the rails 
at the end of each block. Between the pole swinging from 

Fig. 322. Westinghouse Electro-Pneumatic Tunnel Signal, Train Stop, and Relay and Resistance Boxes, District Railway. 




Fig-. 324. Diagram of Connections for \Vestinghouse Automatic Signals on Electric Railways. 

a pivot is an electro-magnet, c, wound to a high resistance 
and connected through a contact operated by the track coil 
armature between the positive rail and the negative main, 
and carrying an arm d with a contact which opens or closes 
the local circuit controlling the signal motors. The relays 
act thus : When a difference of potential exists in the nor- 
mal direction between the rails, i.e., when there is no train 
in the block-section the track coils are excited and draw 
up the armature, e, and thus close the circuit through the 
swinging coil, r, which is attracted to one of the poles of 
the relay and thus closes (through contact on d) the local 
signal current. The two relays A and B are duplicate in 
every respect and operate normally in a precisely similar 
manner a contact in the local signal current, and as these 
currents are in series, unless both are closed no current 
will pass and the signal will remain at danger by gravity. 
The great trouble with automatic signalling particularly 
on railways operated by electricity is the liability of the 
relays to be energised by extraneous currents, so as to give And the train-stop is electrically, see fig. 324, in parallel 

a false "clear" indication. The principal source of ex- 
traneous currents is the main power current, and when the 
running rails are not used for the return current this source 
is abnormal. In the \Yestinghouse system, although it is 
possible to energise either or both relays while the block is 
occupied by a train they are so interlocked that it is im- 
possible for them both to be simultaneously energised in the 
normal direction by extraneous currents. The possible con- 
ditions are : 

1. Normal both relays shunted and no extraneous 

current ; 

2. One relay shunted and one energised normal when 

the signal-circuit is broken at one point ; 

3. One relay shunted and one energised reversely 

when the signal-circuit is broken at two points ; 

4. Both relays energised, one normally and the other 

reversely, when the signal-circuit is broken at 
one point. 

Fig. 325. Interior of Earl's Court East Signal Box, Metropolitan-District Railway. 



Fig. 326. Westinghouse 

with the signal, and therefore controlled directly by the 
signal-circuit through the relays. 

The general rule which forbids a guard to give a " right- 
away " signal to a driver if he be unable to see that the 
signal be " off " obtains on the Met. District R., but as the 
guards are not always able to see the fixed signals, indicators 
having two lenses facing forwards and two facing backwards 
are fixed on the platforms. The upper lenses are green and 
the lower ones yellow, which colour was adopted in prefer- 
ence to red, so that it might not be confused with " stop " 

This installation introduced two decidedly novel features, 
viz., the Illuminated Diagram and the Magazine Train 

The Illuminated Diagram, figs. 325 and 326, is hung 
in a frame in the signal-cabin and shows the lines and junc- 
tions just as the ordinary diagram, which is part of the fur- 
niture of the signal-cabin, does. It is, however, drawn on 

Fig. 327. Transmitter of Westinghouse Magazine Train-Describer. 

glass, which is opaque all over except where the lines are 
drawn. The back of the glass is partitioned off into lengths 
which correspond to the "Track-Circuit" block sections 
and between the partitions are electric lamps connected to 
and controlled by the " Track-Circuit " relays, so that as 
the block-sections arc occupied by a train the lamps, which 
are normally burning, are cut out and the occupied sections 
and fouled junctions become dark. It is surprising to see 
how easily signalmen work complicated stations, e.g., Man- 
sion House, practically in the dark, as they can see but few 
of the points and signals they work, and require no advice 
of trains or block instruments, the positions of the ap- 
proaching and departing trains being indicated on the illu- 
minated diagram. 

Illuminated Diagram. 

The Magazine Train Describer, figs. 327-328, takes the place 
of the ordinary bell-codes or train-describers, which, owing to 
the use of automatic signals, and there being more than one 
block section between adjacent signal cabins, are imprac- 
ticable, because there would be no permanent record of the 
indications, and the signalman would have to trust his 

Fig. 328. Combinator and Receivers of Westinghouse 




memory to set the correct roads and signals required by the 
trains. One length of the Met. District R. has 25 block 
sections between two adjacent signal-cabins. 

The describer consists of two instruments ; the Trans- 
mitter, fig. 327-, in one cabin and the Receiver, fig. 328, in 
the cabin at the other end. The transmitter illustrated pro- 
vides for 15 different indications: on its face are 15 
circles on which are the headlight arrangements of 
the various trains. When a train departs the signalman 
turns a pointer to the circle corresponding to the "marker" 
(which indicates its destination) on the train, and then pulls 
the handle on the right, which movement completes one to 
four circuits connected by line wires to the receiver. The 
four circuits represent the letters A B C D, out of the various 

against the inner end of those studs which have been pushed 
in by the hammers. \Vhen the drum revolves after receiving 
a train description, it carries these springs with it, but the 
springs are also capable of a movement in the opposite direc- 
tion to the drum being imparted to them, one step at a time. 
Each time that the receiving signalman passes a train on 
to the next section a current is picked up from each of 
the studs which have been pressed in by the hammers, and 
by means of the combinator (consisting of four magnets, 
and shown in fig. 328 above the receiver) these currents close 
one of the 15 circuits corresponding to the description sent. 
Each of these circuits actuates a drop of an annunciator 
on a shelf over the locking frame. As long, therefore, as the 
springs rest on studs which have been pushed in, the cor- 

- 3 2 9- Westing-house Train Destination Indicator, showing the Destination of the next three Trains, District Railway. 

combinations of which a code to indicate the various trains 
is made up, and a commutator in the transmitter groups 
the four circuits in 15 different ways, each of which corres- 
ponds to one of the 15 positions of the pointer above referred 

The other ends of the four line wires are connected to 
four electro-magnets, each provided with a hammer which 
strikes one of the studs in the periphery of the drum of the 
receiver (fig. 328). The hammers press the studs in, and 
after the electro-magnets have been energised the drum auto- 
matically turns through a division, and so presents another 
set of studs to the hammers when the signalman at the trans- 
mitter describes the next train. 

Inside the drum is a set of four springs, which wipe 

responding head light appears in the annunciator. Only one 
indication is shown at a time, and that is a description of 
the next train approaching. The man at the transmitter 
sends, as each train leaves, a description of the train. This 
is registered on the magazine, but as only one train can be 
shown on the annunciator at one time, the remaining indica- 
tions are stored up, and after a train has arrived the signal- 
man cancels the indication by means of a plunger which 
moves the springs forward, and the description of the pre- 
vious train disappears and that of the next train appears in 
the annunciator. The springs, in passing from one set of 
studs to the next, automatically release the previous studs, 
so that these are ready to be again operated on by the ham- 
mers. The drums are self-winding, and therefore the studs 
may be utilised over and over again without attention. 



Fig. 330. Tube Railway Signals, 

Fig. 331. Point Motor in Tube Railway. 



Train Destination Indicators, fig. 329, are suspended 
over the platforms with the object of informing passengers 
of the order in which the trains will arrive. This appliance 
consists of glass screens on which are painted the possible 
destinations of the trains, with figures i, 2, 3 beside them. 
These figures are invisible until they are illuminated by the 
incandescent lamps placed behind them. 

The currents for lighting these lamps are controlled by 
a magazine train-describer (fitted with three sets of springs 
instead of one as above described), placed in the circuit be- 
tween the adjacent signal cabins so that the same sequence 
of trains as in the cabin is stored up. \Yhen a train leaves a 
station it automatically cuts out all the lamps behind the 
figures and cuts in another set, so that the i disappears, the 
2 becomes i, the 3 becomes 2, and a new 3 appears. 

'Lerersihonnthus 6 
nit/tout circle 

17. Asking lever for 

Tory road shunt 

Fig. 332. Signals at Coven t Garden and Holborn. 

These Train Destination Indicators are provided at all 
the tunnel (and some other important) stations on the Met. 
District R. At converging junctions, such as Earl's Court, 
where the signalman takes the trains in the order that suits 
best, the indicators are provided with illuminated shafts in- 
stead of figures, and by accepting a train he causes an illu- 
minated arrow head (pointing towards the line on which 
the train will come) to appear at the end of the shaft on the 
indicating screen. 

To prevent the signalman getting " out of step " when 
cancelling the descriptions on his receiver, the plunger is 
locked directly there are no more descriptions on the drum, 
and this locking informs the signalman when the last train 
for the night has passed. 

Each block-section is numbered. The signals and 
" Track-Circuits " have the same numbers prefixed by the 
letters S and T respectively. Signal-cabins are designated 
by letters. 9 

The installation includes 410 " Track-Circuits " and 488 
signals. The number of signal movements per diem is about 
<K>,OOO, with a maximum train service of 33 per hour. 

The air for working the points and signals is conducted 
along the whole length of the line in a pipe 2ins. diam. It 
is compressed to a pressure of 65 to 70 Ibs. per sq. inch 
above the atmosphere. The compressors are electrically 
driven, and are in duplicate at each of the sub-stations. 

Baker Street and Waterloo R. 

Great Northern, Piccadilly and Brompton R. 

Charing Cross, Euston and Hampstead R. 

These are deep level Tube railways worked by electric 
power and signalled by similar appliances to those on the 
District R. 

Owing to the fact that the Tube is only lift. Sjins. 
diam. the signals and point connections had to be modified. 

The type of signal employed is illustrated by fig. 330. It 
will be seen that the frame carrying the spectacle glasses 
(the red being the upper and the green the lower) is coupled 
to a weighted lever. This lever is carried by a pin through 
the casting carrying the lamp. Behind the lamp is the 
signal motor, which, when air is admitted, drives the piston 
to the left, and the end thereof, passing under the weighted 
lever, raises it and, consequently, also the spectacle frame. 
When the signal has to return to the "on " position the air 
is exhausted, and the weight on the lever brings down the 

spectacle frame and forces back the piston. 
l"-** 6 **^ ^^ Under the lamp casting is the indicator 

box, whereby the "return-indication" is 

given in the locking-frame that the signal 
has gone to danger in those cases where the signals are 
operated from signal-boxes. It also releases the corresponding 
distant signal, where such signals are provided. The distant 
signals have yellow glass for the '' on " and green for the 
" off " position. 

Behind the signal is the valve for operating the automatic 

Fig- 331 illustrates how the point motor and its connec- 
tions have to be secured to the side of the tube, so as to leave 
the " four-foot " free. 

At some stations there are small boxes which are 
required in order to work the crossover-roads at those 
places, and these boxes are only open when traffic has 
to be worked through those points. The signals operated 
from all the signal-boxes are automatic, and return to dan- 
ger when a train passes them, independently of the action 
of the signalmen. Such signals are known as semi-auto- 
matic. The signals are spaced so as to allow for the train 
service to be on a li minute "headway." They are pro- 
vided with " long-burning " oil lamps, which burn a week 
without requiring any attention. 

A feature of the signalling on the " Piccadilly Tube " is the 
arrangement for getting the cars from the Strand Branch in 
order to send them to the Lillie Bridge yard every alternate 
night for examination, etc. 

The junction of the Strand Branch at Holborn is arranged 
as shown in fig. 332. The line from Finsbury Park to 
Hammersmith passes under the Strand Branch, and there 
are no points, consequently no facing points, in that line. 
The only connection is with the " eastbound " line, and 
therefore there must be a " wrong-road " movement over 
that line from the junction up to the crossover-road at Covent 
Garden. To safeguard this operation special levers have 

X 2 








5 Signals. 

? Levers fof faints 

. rar.iny PQ'nfl l QtH-i, 


Spare Le vr3 

been provided in both the Holborn and Covent Garden boxes. 
When the man at Holborn wishes to send a train in the 
wrong- direction he pulls his No. 17 lever half over. This 
he can do at any time without interfering with the traffic, 
and it gives an indication in Covent Garden box that the 
movement has to be made. If the man in the latter box 
is in a position to deal with the train he opens his crossover 
road, 6, and pulls off signal 10. He then pulls over his con- 
trol lever, 8, which electrically frees the completion of the 
full stroke of 17 lever at Holborn. When the control has 
been accepted by the lever being pulled fully over, 8 lever at 
Covent Garden is backlocked, and consequently 6 crossover, 
which is interlocked with i and 2 and 1 1 running signals. 
The right-hand signal, A, operated by 19 lever at Holborn, 
is now free, and when lowered allows the wrong-road move- 
ment to be made. The signals are semi-automatic, and are 
restored to normal as the train goes by, but none of the 
levers can be reversed until 17 at Holborn has been put half 
way back, and this cannot be done as long as the line is 


Great Northern and City Railway. 

The Great Northern and City R. runs Jrom under Fins- 
bury Park Station to Moorgate Street. It is electrically oper- 
ated, and is constructed to pass ordinary 
main line carriages, but it has no physical 
connection with the other tube railways 
of London. It is signalled automatically. 

Fig. 333 is a diagram of the line, 
from which it will be seen that there are 
two termini and four intermediate sta- 
tions. Each of these have an outer 
home, inner home and starting signals. 
At Drayton Park there is a siding con- 
nection with the down line, and at the 
terminal stations there are scissors 

~ Styno/s. Automatic Contra/. 

Fig. 333. Diagram of Line ; 

Signal-boxes, with attendant signalmen, are provided at 
each station. These men are not necessary for signalling 
purposes, and they do not interfere at all with the operation 
of the signals, but they are useful for booking the passage 
of the trains and watching that all is right, whilst their pre- 
sence would be exceedingly useful in case of emergency. All 
the relays and operating mechanism, too, are in the signal- 
boxes, so that the men keep them under constant observa- 

The signals have no moving parts or any mechanism at 
all. They consist of a lamp with two lenses in front, an 
upper red and a lower green, and behind each of these is 
an incandescent lamp. Only one of these can be burning 
at the same time, and it is the illumination of the red or the 
green lense that indicates whether the section be clear or 

"Track-Circuits" are employed, but in order to guard 
against extraneous currents interfering with the operation ol 
the signals an additional electrical contact, which comes into 
operation by the passage of the train itself, is introduced. 
This contact is in the form of a treadle, fixed from 350 to 
40oft. in advance of the^signals, and comes in contact with a 





Fig. 334. Diagram of Electrical Connections ; 







* Automatic Signals 

P/unger Control Stgnols. 

Great Northern and City Railway. Fig. 333. 

brush on the last vehicle. It is therefore guaranteed that 
the whole of the train has passed, and the distance of 350 
or 4ooft. acts as an " overlap." The treadle contacts are in 
circuit with the 500 volt operating current and a release coil 
solenoid fixed on the mechanism of each signal in the signal- 
box. Consequently any extraneous current of less voltage 
than 500 volts has no effect upon the release coils. The 
current for the track relays is supplied from one of two 100 
volt motor generators fixed at Highbury Station. The pres- 
sure is reduced by lamp resistances to about 3 or 4 volts- 
Fig. 334 is a diagram of the electrical connections. 
The train is assumed to be on the down line between Old 
Street and Essex Road, at the point where the arrow is 
shown. This short-circuits the track between insulated 
joints 2.3., causing the armature of relay i to fall and de- 
energise solenoid 5 in Old Street box so that the spindle 50 
falls and contacts G.G. are broken and R.R.C. made. This 
switches off the green lamp of Signal S and opens the red 
and closes the circuit to treadle 7. 

When the spindle 511 falls the armature 6 comes to rest on 
the top of it and locks it down so that should the relay i 
become energised from any improper cause the signal cannot 
turn to green again until the right moment 

When the train passes out of the section 2.3 
and enters 3.4. relay 8 is de-energised and the 
spindle of solenoid 9 falls, so that Essex Road 
outer home signal is changed from green to red 
and locked there. Signal S still remains locked 
though, even if the train has passed out of sec- 
tion 2.3. and relay i is energised. The lock is 
not taken off until the brush on the last vehicle 
of the train strikes treadle 10, which completes 
the 500 volt circuit to the coils of the armature 

Signal f.Aufoma*c 

6, attracting the latter and allowing the spindle 50 to rise 
and to switch signal 5 from red to green. 

Should the 100 volt current fail all the solenoids would 
be de-energised, so that the spindles would fall and the 
signals go to red. Should anything go wrong with the 500 
volt current the treadle would not attract the armature coils 
and could not release the spindle. All failures are therefore 
on the side of safety. 

Pilot lights are provided in the signal-box to indicate that 
the lights are burning, and correct signals being given. They 
also indicate the approach, departure and positions of trains. 

At the terminal stations the points and signals are worked 
by levers the points by rodding and the signals electrically. 
Fig. 33;; is a diagram of the electrical interlocking connec- 
tions at Finsbury Park, and they are the same at Moorgate 

There is one outer down home signal and splitting inner 
home signals. All these are worked by one lever, the position 
of the points selecting the inner home. 

' Track-Circuits " are provided throughout the whole 
length of the platform roads, so that the inner home signals 

Great Northern and City Railway. Fig. 334. 


cannot be changed from red to green unless the line be clear. 
The outer home signal can be cleared when the previous 
train, by the brush on its last vehicle, completes the circuit 
through the treadles in the platform roads immediately past 
the crossover-road points. 

The lowering of the up starting signals for leaving Fins- 
bury Park are dependent on the previous train making con- 
tact with the treadle in advance of Drayton Park outer up 
home signal. 

The facing-point locks on the crossovers are locked in 
when set, and remain so until the train has passed over. 

The patents for this system were taken out by Mr. R. P. 

lulls station there are some power-worked (electrical) stop 
and distant signals on the up and down west main-lines. 
Each stop signal has a corresponding distant, and electrical 
treadles are provided to indicate when the whole of the train 
has passed. There are also similar signals near Ainsdale. 

Metropolitan Railway of Paris. 

When this line was first opened it was signalled by the 
Cie. de Signaux Electriques pour Chemins de fer, Paris, with 
automatic signals on the Hall system. 

' Track-Circuits " were not employed and the signals 
were lowered through contacts made by electrical treadles. 

Clears Oroyf-on Park Adrdnct Sect/on Block 

!'' 335- Diagram of Signal Connections, Kinsbury 

Brousson (the engineer and traffic manager of the line) and 
Mr. A. H. Binyon, of Spagnoletti and Co., who carried out 
the installation. 

Liverpool Exchange. 

The Lancashire and Yorkshire R. have put down some 
' Track-Circuits," automatic signals and locks on signal 
levers on a portion of their electrically operated Liverpool- 
Southport line, near Liverpool. 

At Liverpool (Exchange) station " Track-Circuit " is laid 
down in Nos. 9, 10 and n roads. Scissor crossings are pro- 
vided between Nos. 9 and 10 roads, the points of which are 
mechanically operated, but they are electrically detected. The 
eight signals applicable to these crossings are actuated by 
Sykes' motors, the current being set up by the movement of 
the usual lever in the locking frame. Under the home signals 
at the entrance to the station are lower distant arms which 
" follow " the signals at the scissor crossings and indicate 
whether the line be clear or not in a similar way to that sug- 
gested on p. 24. Brown's relays (fig. 324) and Sykes' in- 
sulated joints (fig. 146) are provided for the " Track- 
Circuits," and an illuminated diagram showing these lines is 
fixed in the signal-box. 

Between Liverpool Exchange Station Junction and Sand- 

100 Volt Mams from Highbury. 
Park Station ; Great Northern and City Railway. 

Each train was protected by two stop signals which have 
red and white lights. Originally these signals were on the 
" Normal-clear " method, but they were subsequently 
changed to the "Normal-danger." 

The working is as follows : Let A B C D be four signals, 
each of which has a treadle deflected by the rail which 
actuates relays on the signals. A train travelling from A to 
D has passed signal C, and when the treadle is deflected, 
signal C is thrown to danger (A and B being already at dan- 
ger), and signal D lowered, providing the preceding train 
has passed over the two next treadles at E and F. At the 
same time a relay at A is freed, so that the signal can be 
lowered when a second train enters the section immediately 
behind A. The deflection of a treadle throws its own signal 
to danger, lowers the signal in advance of it if two sections 
in advance be clear, and frees the second signal behind it. 
Each train is, therefore, protected by two stop signals, and 
before a signal can be lowered the preceding train must have 
gone through the next two sections. 

The signals consist of a cast-iron box with two openings 
an upper one with a red light, and a lower with a white. In 
the case is a disc operated by an electro-magnet. The disc 
is weighted so as to normally blind the white light, and clear 
the red. XYhen the electro-magnet is energised the disc is 
raised, the red light blinded and a white signal given. 



That the work is effectively done is proved by the follow- 
ing figures, which are really wonderful : The train service 
commences at 5.30 a.m. and ceases at 12.30 a.m., there being 
a two-minute service of trains. Each wheel deflects the 
treadle, and there are 103 treadles on this part of the line. 
It is estimated that from July I5th, 1002, to January, 
1903 (the time of the Author's visit), about 5,000,000 axles 

had passed over each treadle, or a total of 515,000,000 deflec- 
tions, which operated upwards of 64,000,000 signal move- 
ments, which had all been performed without a single failure, 
nor was there any apparent wear and tear. Owing to the 
frequent service the maintenance work is performed with 
great difficulty, and yet only three men are required to look 
after 103 signals. 

Lighhnq Mctns 

Fc 1 














*/ / 

, J 







r . 



> o 


o o 





355- Diagram of Signal Conneciions, Fin-bury Park Station ; Great Northern and City Railway. 



pig.. 33 6. Automatic Signalling, Metropolitan Railway of Paris. 

The recent extensions of the Metropolitan R. of Paris 
have been signalled, automatically, on a principle designed 
by some of the officers of the railway. It is illustrated by 
fig- 33- The system is similar to that in use on the Great 
Northern and City R. see figs. 333-4. 

Solenoids are employed to operate the signals, and a 
treadle is used as in the Great Northern and City. Each 
train is protected by two stop signals. The arrangements 
on the extreme right are those at a terminus. Alternating 
current is used and the signals are "normal-clear." Auto- 
matic stops (controlcurs de passage a 1'arret) are employed. 

After this work had been completed it was pointed out 
that should a shoe get damaged and the treadle not be 
operated, the signals behind it would not go to danger. 
There was also a chance of a short-circuit, which might lead 
to only one signal going to danger and a second " short " 
might lead to no signal protecting the train. As a conse- 
quence the Cie. de Signaux Electriques put in new connec- 
tions as shown in fig. 337, and the signals were also altered 
to the " normal-danger " system. 

A train is assumed to be travelling from right to left and 
to be approaching post A- When it passes A the brushes on 
the train make contact with the treadle, which causes the 
following circuit to be completed : Wire 16, wire 17, lock 
L, wire 18, wire 19, wire 7, and negative rail. This causes 
the lock L to attract the armature L 2 so that the contact 
carrier L 3 falls and contact is made at 3, 4 and 7, 8 and n, 
12, and broken at I, 2 and 5, 6 and 9, 10. 

Previous to the arrival of the train a white light had been 
shown at A, and this has been made by the following circuit : 
The circuits are " normal-danger," and so an approaching 
train causes a current to come from the post in the rear of A 
by wire i, to coils I of the annunciator X, wire 2, wire 3, wire 
4, wire 5, wire 6, wire 7, and negative rail", so that the coils 
i attract the armature of the bell so that contacts x y are 
joined. The commutator V is now energised by the following 
circuit : Positive rail at Post B, wire 8, wire 9, contacts 5 
and 6, wire 10, wire 1 1 up to post A, resistance R, wire 38, 
contacts i and 2, wire 12, contacts x and y (joined by the 
armature of the bell when coils i were excited), wire 13, to 
commutator V, wire 14, wire 15, wire 2 and wire 3, wire 5, 
wire 6, wire 7 to negative rail. This current energises com- 
mutator F and so relay F 2 is attracted, and this completes a 
circuit to the white lamp as follows : Positive rail, wire 8, 
wire 26, contacts p and m, wire 30, white lamp, wire 31, wire 

15, wire 2, wire 3, wire 4, wire 5, wire 6, wire 7, and negative 


When the train passes A and the contact carrier L 3 falls 

contacts i 2 are broken, and the circuit to the commutator 
F so that the relay F 2 falls, and circuit p m to the white light 
is also broken, but that p n to the red light is made. 

Contacts 5 and 6 join up the signal wire to the positive 
rail. The separation of these contacts and the joining up of 
contacts 7 and 8 put the signal wire on to the negative rail'. 

The contacts 9, TO and n, 12 are very important. The 

Fig. 337. Signalling, Metropolitan R., Paris. 

former, when joined, complete the circuit to the relay S, which 
raises the contact carrier L 3 . When they are broken and 11, 
12 made the Releasing Wire is joined to the treadle by this 
circuit: Treadle, wire 16, wire 20, contacts n and 12, wire 
21, wire 22, to relay S at post in the rear so that the latter is 
excited and the contact carrier L 3 at the post in the rear raised 
when a train passes the post in advance. 

From the same contacts- n and 12 current flows thus: 
Wire 21, wire 23, resistance, wire 24, coils 2, wire 25, 
Bell and Controller Wire to post in advance, and then at post 
in advance wire i, coils i, wire 2, wire 3, wire 4, wire 5, wire 
6, wire 7, and negative rail. Coils 2 at the post in the rear 
are energised so that the bell A' is attracted and contacts x 
and y are broken, and therefore the armature F is de- 
energised and relay V- falls, if it have not already done so 
when contacts i and 2 were broken. 



THERE is a decided want for an effective system of sig- 
nalling for electric tramcars on single lines, in order to 
prevent a car leaving a passing place and entering a section 
when another car is approaching from the opposite direction 
over the same section; but very little has been done in this 
matter in this country. 

In America, where there are high-speed interurban 
trolley-car lines over which cars travel at as high a rate of 
speed as steam trains, more attention has been paid to the 

For single tracks the system should be so arranged that 
the signal at the far end must be at danger before the one 
at the near end is lowered. If only one car is to be in a 
section at one. time then neither signal must come off so long 
as a car is in the section. If any number of cars travelling 
in the same direction are allowed on the section, then the 

Fig. 3.41. Dover Tramways. 

signal at the far end must not fall until all the cars have 
arrived from the opposite direction, and the system must 
be so arranged that each car gets the signal to enter the 
section. The signal should take the form of an ordinary 
semaphore arm or disc for day, with spectacles or an arrange- 
ment of lamps for night use. The system should be electric 
and not mechanical, because with the latter system the wires, 
etc., might be interfered with by mischievous persons. 

In this country, at Dover, there has been working since 
1899 a system of block for single-line tramways, and which 
is the invention of Mr. H. I. Stilgoe, now City Surveyor and 
Engineer, Birmingham, and is manufactured by Saxby 
and Farmer, Ltd. It is illustrated by figs. 341 and 342. 

At the crossing places a pillar is erected on the foot- 
path carrying an instrument which has a miniature arm and 
a black and a white plunger. The arm when up indicates 

that the section ahead is occupied, and when down that it is 
clear. Before the car enters on the section the driver must 
press in the black plunger, which will throw up the miniature 
arm above the plunger and the corresponding arm in an 
instrument at the other end of the section. When he leaves 


f Magnets 

Commutator ox Commuiafor on 

e*4of fit* Plunger end of White Plvnfc 


the section he must depress the white plunger, which will 
again lower the miniature arms. 

The idea is no doubt a good one and may be adaptable 
to the requirements of some towns and cities, but it would 
not do for a high speed trolley-car service where signals are 
the more needed. 

Hanson Signal. 

Mr. Harison Jones has designed a system of signalling 




lor electric cars on single tracks which is in use at Southsca, 
Twickenham and Swindon. 

Near the outlet from the double line or crossing place a 
signal arm is attached to one of the pillars carrying the 
trolley wires. In the trolley wire itself is a switch which is 
opened by the pulley, and this sends a current locking the 
signal at the other end of the section. 

The arrangements are illustrated by fig. 3-13. 

A car approaching from the end A makes contact by the 
overhead switch i, and a current is allowed to flow through 
the signal Si, preference solenoid p, by the auxiliary wire 
and through the locking solenoid \2 of the signal 2 at the far 
end of the section, thereby locking the signal 82 at danger. 
If the line be clear, the arm of Si will be lowered showing 
a green light at night. The car passes off the contact i and 
breaks the circuit to the signal, which returns to danger as 
the arm is counterweightcd to do so, and then the car reaches 
switch 2. Here a current is allowed to flow through the 
locking coil li of the signal Si, which locks this signal at 

Both signals arc now locked at danger and the car passes 
over the section protected until it reaches the loop at B, when 
it makes contact on switch 3 and passes a current through 
the two unlocking solenoids U2 of 82 and MI of Si in scries, 

unlocks both signals, leaving them free to be worked by a 
car approaching in either direction. 

There are cases in which the two lines of a double track 
foul each other, and where it is not desirable that cars should 
pass each other on the narrow portion. The Harison auto- 
matic signal provides for this contingency by the arrange- 
ments shown on fig. 344. 

A car coming in the direction shown by the arrow passes 
off the main section on to section i. Here it draws its 

k Length of Section tote protected *f 


/ Serf ion Insulator 
Length to suit requirements 
usually 40 'yards 

current through the Signal Si in series with the preference 
solenoid or coil p l , section 3, locking ceil 1- of .S' 2 , thereby 
locking this signal by the tap to the main line, and if the line 
be clear the signal arm of Si will be lowered, showing a 
green light at night. The car then passes on to section 2 ; 
here it draws its current through the locking coil li of the 
signal Si, by the tap to the main line, thereby locking this 
signal. Both signals a-re now locked at danger so that the 
cars are prevented from entering the section at both ends. 

This Contact must be abreast 
of the Light marked A 

ig- 345- Diagram of Electrical Connections, Brecknell, Munro and Rogers* System. 



The car passes on to section 3 and travels over the pro- 
tected section. If it is not essential that the car should 
draw current on this section, as for instance on a down 
gradient, it can do so, if necessary, through the locking coil 
\2 of the signal 82, by the tap on the main line. 

On entering section 4 it draws its current through the 
unlocking 2 of $2 by one of the auxiliary wires in series 
with the coil ui of Si, by the tap to the main line, thereby 
unlocking both signals. 

Brecknell, Munro and Rogers' System. 
The Burton and Ashby Light Railway (which is the pro- 
perty of the Midland R.), the Wemyss Light Railway, and 
certain portions of the electric (overhead trolley wire) tram- 
ways at Bristol, Bath and Sheffield, have been equipped by 
Messrs. Brecknell, Munro and Rogers, Ltd., electrical en- 
gineers, Bristol, with the automatic signals illustrated and 
described below. 

Fig. 345 is a diagram of the electrical connections. At 
each end of the section to be protected is a box a, a 2 , con- 
taining, in the upper part a lamp and in the lower part two 
magnets and a rotary switch which, on being turned, causes 
the lamp to be switched in. The lamps are provided with 
an efficient hood so that the sun cannot shine on the lense 
of the lamp. 

At each end there are two contact makers suspended 
from the trolley wire. These are struck by the trolley head 

and make and break contact. For a car travelling from left 
to right the contact d, when struck, causes the magnet e 5 
to be energised so that the rotary switch c turns and switches 
in the lamp f- in box a. When the car passes off the section 
the contact d? causes magnet e* to be energised so that the 
rotary switch is reversed and the lamp is switched out. 
Should, however, a second car enter from the same end 
before the first has passed off the section the magnet e* is 
again energised, and the rotary switch turns another notch. 
The first car on leaving turns it back one notch, but as it is 
not yet normal the light remains in until the second car has 
passed out. 

Contact d* makes contact for the opposite direction and 
energises magnet c 3 , and contact d- breaks it by energising 
magnet e 2 . 

To provide for such a contingency as two cars approach- 
ing a block simultaneously from opposite ends it is important 
that one making or " On " contact should be set 40 yards 
back from the signal lamp or in other words given a lead 
of 40 yards over the " On " contact at the opposite end of 
the block, which should be abreast of the signal lamp situated 
at that end. By this means the simple rule that no car must 
pass a lamp which is lighted effectually prevents cars meeting 
in a section. Should two cars come under the contacts at 
each end simultaneously one motor man would be standing 
a trolley pole's length in front of his danger signal due to 
the angle at which the trolley arm lies back over the car 

First Contact 

Plan showing Box in 

Fig- 347- 

Tierney and Malone Eleclric Point Control (Single Box). 



and therefore cannot see it, and the other motorman is 40 
yards back from his danger signal and has that distance in 
which to pull up. By this arrangement of lights and contacts 
it is not possible, should the cars have passed through the 
contacts at precisely the same moment, for a danger signal 
to be visible to the motorman at each end of the section at 
the same time, one motorman only sees a danger signal. 

Experiments have proved that should a second car enter 
the section just as the first was leaving and making and 
breaking contacts made together a very remote contingency 
the lamp would remain in, indicating that another car had 
entered the block. 

Working Tramway Points Automatically by Electric Power. 

Messrs. Brecknell, Munro and Rogers, Ltd., arc also 
the makers of the system of electrically operating tramway 
points invented by Messrs. Tierncy and Malone, and which 
is illustrated by figs. 346 and 347. 

By this arrangement the switches of a junction are 
automatically operated. Let into the ground, and with the 
top level with the same, is a cast iron box a which contains 
two solenoids M 1 M 2 which are alternately energised, and the 
points m m 2 of these bind against a way-beam, on the upper 
part of which is an arm b coupled to the switch c, so that the 
alternate energisation of the solenoids opens and closes the 

The first. or scries solenoid is energised by a circuit set up 
by the car making a contact with the first contact C 1 , which 
is fixed si yards to the rear of the switches and insulated from 
the battery wire. This causes a current to pass to the car 
through the solenoid M 2 . The current from trolley wire to 
earth is as follows : Line tapping T, fuse F 1 , switch S l , 
solenoid M 2 , back to switch S 2 , fuse F 2 to contact C 1 , and 
thence to trolley head on the car and so to car motors and 
to rail return. It is obvious that the current can only flow 
along this circuit when the car is taking power from the 
line. The track point or switch can therefore be operated by 
the motorman keeping his controller on when going over 
contact C 1 . Or it may not be operated, if it already lie in the 
desired position, by the motorman having his controller off 
when he passes over contact C 1 . In the latter case no cur- 
rent will flow to the solenoid M 2 , as the circuit will be open 
at contact C 1 . There is also a shunt circuit to earth through 
lamps D 1 D 2 , which act as non-inductive resistance for taking 
discharge of magnet coil. 

The green signal lamp G is permanently connected to 
line through switch fuse S 4 , and an alternative circuit to 
earth is provided through red lamp R 1 or red lamp R 2 , the 
selection being made by the tongue of the track point accord- 
ing to its position- whether set for the straight line or the 
branch line. The lamps are earthed by the way-beam coming 
in contact with the terminals. 

These lamps indicate which way the points lie according 
as to whether the right or left hand lamp is switched in. 
The green lamp is always burning. Should it be out, then 
the motorman is warned that something is amiss. In ad- 
vance of the points is the second contact whereby the other 
solenoid M 1 is energised so that the switches arc restored. 

This solenoid is in shunt with the trolley wire and is 
earthed direct. 

Electric Tramway Equipment Co.'s Signal. 

Fig. 348 gives details of the apparatus fixed by J. G. 
White, Ltd., on the Swindon tramway, and of which the 
Electric Tramway Equipment Co., of Birmingham, are the 

On a pillar, about 6ft. Gins, above ground level, is fixed, 
near the entrance to each end of a single line section, the 

-TT. 7- --Hi 


Fig. 348. Signal, Swindon Tramways. 

cast iron box a which contains a two-way switch lever b 
and two lamps c c. Green glass is provided in the sides of 
the box a. 

As long as the lamps are burning no car may enter the 
section, but, if they are out. a car may enter after the con- 
ductor has turned the switch which switches in the lamps at 
both ends. When the car leaves the conductor reverses the 

Siemens Bros.' Signal. 

This is in use on the Pontypridd Tramways, and is illus- 
trated by fig. "549, which shows two cars C 1 C 2 . The former 
i* in possession of the section and is travelling from right to 

Two lamps arc fixed at each end of the section, L 1 (green) 
and L 4 (red) at one end and L 2 (red) and L 3 (green) at the 
other. When car C 1 passes the section insulator I 1 it took 
current from the feeder F through the solenoid S 1 (fixed in 
a case by the side of the track) so attracting armature A 1 and 



completing the circuit for the lamps L 1 and L 2 from the 
feeder F to the earth. A green light was therefore shown to 
the motorman of car C 1 and a red light was shown in the 
opposite direction towards car C 2 at the other end. These 
lamps are fixed 30 or 40 yards in advance of the section 

" 7 * 

Fig. 349. Siemens Bros.' System. 

When the car C 1 has gone through the section it passes 
under section insulator I 2 and the armature A 1 returns to 
its normal position and the current is taken through solenoid 
S 2 and lamps L 1 L 2 are switched out. 

Lamps L 1 L 2 and L 3 L 4 are in series, so that a green light 
at one end ensures a corresponding red light at the other. 
Xo car must enter a section against a red light nor if the 
green is not given. 

Where lines cross each other, as in fig. 350, the passage 
of a car from D to C, under the section insulator in this 

case M 1 causes current to be taken from the feeder at the 
crossing, so energising solenoid S 1 and attracting the arma- 
ture and switching in the four lamps I I F F. Green or white 
lights are shown in the former two and red in the latter two. 
Having passed the crossing K current is taken from the 
other side of the section insulator at K, so magnetising the 
other solenoid S 2 and putting out the lights. 

Siemens-Hahke System. 

Two green signal lamps g h, fig. 351, are employed in 
this system and two red, / i, also two contacts b c at one end 
and two, d c, at the other, which respectively are connected 
to solenoids fe /. These solenoids operate ratchets q r which, 
when the section is clear, are in the dotted position and clear 
of the teeth on the disc n. Should a car enter from the 
right, contact would be made between d e, which would ener- 
gise solenoid / and so attract the ratchet r and draw it to- 

wards the solenoid so that the disc ti is turned one notch 
and the segment p joins the contacts t v so that a circuit is 
completed to the green lamp g and the red lamp /. When 
the car leaves at the other end the ratchet q is attracted by 
solenoid fe owing to contacts b c, so that the disc n revolves 
backwards for one notch and contact is broken between t v 
and the lamps are switched out. 

Fig. 351. Siemens and Halske System. 

A second car entering on the right before the first leaves 
moves the disc one more notch forward and so on, and for 
every notch forward there has to be a backward movement 
as each car leaves. 

Cars entering on the left energise solenoid fe through 
contacts b c and contacts s u are joined by segment o so that 
green lamp h and red lamp i are switched in. 

Should a car enter at each end simultaneously, and the 
four contacts b c d e were made together, the car from the 
left would get right of way, as solenoid fe is more powerful 
than I. 

U.S. Electric Signal Co.'s Signal. 

In Col. Vorke's report on his visit to America in the 
autumn of 1902 he refers as follows to the use of signals on 
the street railways of Pittsburgh : 

" In places where a double track joins a single track, and where the 
view is bad, the lines are filled with automatic electric signals of a 
simple description operated by the trolley, which inform the motorman 
of a car whether the single line ahead is occupied by a car approaching 
from the opposite direction or not. This is a useful arrangement and 
makes for safety." 

By the courtesy of Mr. Uhlenhaut, the chief engineer of 
the Railways Co., the Author is enabled to give a description 
of the system of the United States Electric Signal Co., of 
West Newton, Mass., U.S.A., whose signals are used on the 
Pittsburgh and many other electric tramways. 

The object of the system is to protect either one car or a 
predetermined number of cars going in one direction against 
cars coming in an opposite direction, also against cars 
following the operating car, also to indicate to each following 
car that it has operated the signalling device. It also secures 
the locking of the signals at both ends of the block, whereby 
the opposing cars cannot affect the signals set against them, 
the release of the said signals resting solely with the cars that 
set them. 

It is also arranged that upon the simultaneous arrival of 
two cars, one at each end of the block respectively, going in 
opposite directions, the car going in one predetermined 



direction shall always secure the right of way over opposing 
cars. Further, in the case of interruption of power, the 
signals go to danger. The signals are cleared and restored 
without slowing up or moderating the speed. 

One of the special features is the fact that all of the fore- 
going is accomplished with one wire circuit. All electric 
contacts are in oil and wearing parts arc extra strong, so 
that there are few repairs. 

Attached to the trolley wire and supported by a span wire 
is an automatic switch. This switch is designed to operate 
one set of signals when the car goes in one direction and 
another set when going in the other direction, lighting the 
signals on the section in front of the car and extinguishing 
the signals on that in the rear. A switch will operate by the 
shallowest or the deepest grooved trolley wheel of any 
standard make, and is neatly encased in metal and thoroughly 
protected from the weather. 

The signals are usually placed a short distance beyond 
a turn out, i.e., inside the single line section, and consist of 
substantial cast iron cases. The doors of these cases are 

I r-f-ww/rf LAMP 


I'.S. Electric Co. Signal. 

provided with upper and lower windows, the upper being 
covered by a red glass and the lower by a white glass, two 
no volt lamps are in each box, one just behind each window. 
Two wires connpct this box with the corresponding one at the 
other end of the section, and No. 12 galvanised iron wire, 
covered with a weather proof braid, is well adapted to this 
purpose. The further box is also connected with the trolley 
wire through a switch similar to the one before described. 
Each signal box contains three operating magnets, and the 
power developed is much greater than is required to operate 
the signals, and thus a very great margin of safety is 
obtained. The three armatures are the only moving parts in 
the boxes. The armature of the lighting magnet is mechanic- 
ally locked in place after it has operated the proper contact. 
This armature is also held by the magnetism and prevents a 
signal from being turned off should the unlocking circuit 
become charged from other wires along the line, and conse- 
quently, if the lamps are extinguished through a failure of 
current, they will immediately be relighted when the current 
returns. They can only be permanently extinguished by the 
action of the putting-out magnet. 

Fig. 352 is a diagram of the electrical connections. 

A car represented by the wheel a is seen in the section, 
travelling from left to right. When this car entered the 
section it closed the right-hand contacts of the trolley switch 
b for an instant, allowing the current to flow over the circuit 
represented by heavy dashes, through magnet c, over line- 
wire 3, through the other signal, to the ground. The path 
taken is shown by the heavy full line. 

When magnet c is energised it throws over its control 
lever, disconnecting the ground at this, the setting end, and 
cutting in a permanent feed from the trolley wire to take the 
place of the switch contact, which opens immediately after 
the car passes. This permanent feed also throws the green 
lamp and disc into the signalling circuit and this indicates 
that a red signal is being exhibited at the opposite end. 

The other set of contacts closed by this magnet complete 
a circuit which starts in the outside contacts of both trolley 
switches. The signalling circuit leads through magnet d 2 at 
the other end, opening a pair of contacts known as the "non- 
interference" device. These contacts open the setting 
circuit from the trolley switch and prevent a car trying to 
enter from the opposite end, locking up the lever to magnet 
c- which would connect both ends of the signalling circuit to 
the trolley wire, making a dead signal until a car passes out 
of the block. 

The circuit indicated by the light broken line is known as 
the releasing circuit. When a car leaves the section, going 
in the direction shown by the wheel, it closes the right-hand 
contacts in the right-hand switch, thus allowing current to 
flow through magnet e 2 which breaks the main signalling 
circuit and also through magnet d which unlocks the lever 
of magnet r. The magnet c now being de-energised and 
the lock open, allows the lever to fall back and the system is 
again in its normal state with no car in the section. 

Eureka Signal. 

The Author was very favourably impressed at the ex- 
hibition at the Street Railway Convention held at Phila- 
delphia, September, 1905, by the Eureka Automatic Electric 
Signal Co., of Lansford, Penn. 

This allows for any number of cars to pass through a 
section in the same direction, all of which are recorded, and 
the signals which were put to clanger by the first car can only 
be cleared by the last car. 

The contact-maker is the chief feature of this system. It 
is suspended above the track immediately inside the turn-out 
and it is struck by the trolley wheel. 

The contact plates arc made of steel combs with teeth 
sufficiently flexible as to permit the trolley wheel to accom- 
plish its work without jump or jar whilst running at the 
highest speed. 

In the two-wire system (the better one) there arc pro- 
vided a double lamp outside the single track, i.e., before 
the turn-out is reached, and which has a red light towards 
the motorman and a green light at its back, i.e., pointing 
in the opposite direction. Immediately inside the single 
line are two lamps above one another. At the other end of 
the section there are similar lamps, but for the opposite 

S 7 G A* .4 L .<? FOR ELECTRIC T R A ^f W AYS. 


Normally all the lights are out, and when such is the 
case a motorman knows that the section is clear. If the 
outer signal, i.e., the single one before reaching the turn-out, 
is at red, he must not enter, but if no light be shown he may 
enter on the single line, and when he passes the contact 
maker one of the green lamps, immediately inside the single 
track is switched in, and both lights in the red and green 
signal at the opposite end of the section. A following car 
may enter if no red light be shown and the motorman will 
know by the green light in front of him that there is a car 
in the section in advance of him and going in the same 
direction. The passage of the second car under the contact 
maker will record itself on the controller, and at the same 
time will switch out the green light in advance and switch 
in its companions, but the far green remains unaltered. The 
former is an indication that all is working properly. Every 
car that follows records itself on the controller, and as each 
passes out and makes contact on the contact maker at the other 


Blake Signal 

Fig- 354 

end, the recorder works backward until all the cars have 
been accounted for. Then the red and green lights are 
both switched out. 

Should a car not go through the section, but reverse and 
come out at the end it went in, it switches out the red light 
at the entrance. 

If a car enters a section against a red light it switches 
out all the lights on the section, which not only attracts the 
attention of the erring motorman, but of all the men running 
in the section. Should the man, on finding out his mistake, 
back out, he switches in the red light against himself. 

Intermediate green lights may be provided to keep the 
motorman advised that all is right. 

Blake Signal. 

The Blake Signal and Manufacturing Co., of Boston, 
Mass., have a very good system, with which the Author was 
greatly pleased. 

It is in use on the Boston and Worcester Electric R. and 
has recently been extensively adopted by the Illinois Traction 
System amongst others. 

Its purpose appeals to those companies who control their 
single tracks by a train dispatcher. Up-to-date lines equip 

their roads with telephones at all calling and passing places, 
and each car must stop at the telephone booth, ring up the 
train dispatcher and ask for orders. This, of course, takes 
time, and there is an unnecessary amount of time spent in 
purely formal communications. 

In the Blake system it is arranged that only when re- 
quired shall the crew be called to the telephone, and they 
may run by all other points unless they have orders to call 
for instructions or wish to do so. 

It should be understood that the Blake signal is not an 
automatic signal and is not so offered. Its description is 
included here for convenience. It is known technically as a 
selective signal whereby the train dispatcher can set a signal 
at any point under his control and stop a car for orders. 
Such a signal is illustrated by figs. 353-354, which shows, 

... v- j-j 

Fig. 355. Blake Switch Board. 

in the former, the signal normal and in the latter the arm in 
the position indicating that the car is to stop for orders. 

The arm is 3ft. 6ins. in length, and it is carried on a 
bracket attached to a trolley wire pole. When the arm is 
down a red light is shown by means of an incandescent lamp 
shining through a red lense. This lamp is normally out and 
is switched in when the arm falls. Should this lamp burn 
out, the upper one is automatically switched in. This second 
lamp is in an interrupted circuit and gives a flashing light 
so that the crew may see that the other lamp has failed and 
report it. This, it is claimed, removes the necessity of 
having a daily inspection of all lamps, as well as from danger 
from a new lamp being defective and burning out shortly 
after it had been put in. 

1 fig- 355 > s a view of the desk in a train dispatcher's 
office. This desk is 3ft. ains. high, 3ft. 6ins. long, and 2ft. 
ains. wide. For each signal there is a pendulum in the case; 
also on each signal is a magnet and corresponding pendulum. 
These vary in length, and those in the desk have platinum 
contacts which are made and broken by the swing of the 
pendulum at intervals varying in length according to the 
length of the pendulum. 



The remainder of the description is best given in the fol- 
lowing words of the patentee : 

The basic principle of the signals is that the time of vibration of a 
pendulum varies with its length. Therefore, if we, at some central 
point, such as a dispatcher's office, start in vibration a pendulum of a 
certain length, allowing it, as it vibrates, to open and close an electric 
circuit, electric impulses will be set up in that circuit, synchronous with 
the vibrating pendulum. These impulses energize the electro magnets of 
all the signals on the circuit, and start vibrations in all of the signal 
pendulums. On only one signal, however, namely, that one whose pen- 
dulum is of the same length and therefore synchronous with the pen- 
dulum vibrating in the dispatcher's office, will these electro magnetic 
impulses be of a proper period to be cumulative in effect, and swing the 
pendulum in an ever increasing arc of vibration. On all the other 
signals, the impulses being out of beat, the pendulums will receive a 
check before they have swung through any considerable arc. On the 
signal which it is desired to set and on the pendulum of which these 
synchronous impulses are accumulating energy, the pendulum having 
reached a certain arc of vibration, mechanically trips a lock, leleasing a 
three-foot semaphore arm which falls to the horizontal position by 
gravity. Having reached a horizontal position, this semaphore closes 
the local signal lamp circuit and also interruptedly closes a shunt to 
ground on the signal line, causing a sounder to draw up in the dis- 
patcher's office and notify him of the fact that the signal has operated. 

For instance, when the dispatcher wishes to set signal No. 9 he 
inserts a plug, similar to a telephone plug, in hole No. 9. This releases 
pendulum No. o,, and also connects the line with the 500 volt circuit 
which is brought to the dispatcher's desk. As the pendulum swings it 
.opens and closes the signal line circuit, sending impulses over the line 
synchronous with its vibration. At the end of 13 seconds this line signal 
pendulum swings through an arc wide enough to trip a lock and drop 
the semaphore arm. The car arrives at the station and the motorman 
or conductor calls the dispatcher by telephone, saying, " Brown and 
Hayes at No. 9 for orders." The dispatcher then transmits his orders 
to the conductor or motorman. The conductor or motorman, if so 
instructed by the dispatcher, then pulls a cord which sets the semaphore 
at " clear " position roady for future operation. 

It should be remarked that until the arm has reached an 
angle of about 45 it is a physical impossibility for the dis- 
patcher to get this indication, and the danger of a false indi- 
cation is eliminated. There are no electrical contacts in 

series with the operating magnets at the various signals. 
The widely varying voltage of the trolley line is met by relays 
which draw up at different voltages, and which may vary 
from 300 to 700 volts. The line consists of a single No. 10 
bare galvanised iron wire on glass insulators. 

Signalling jor High-speed Interurban Railways. 

The Philadelphia and Western R., on which trolley-cars 
are run at high speeds, was opened early in 1907. It is one 
of the very few of such lines that are protected by a proper 
system of signals which in this case was imperative on 
account of the high speed, the heavy gradients and the density 
of the traffic. It is a double line 1 1| miles long. 

The system of " Track-Circuits," protection against inter- 
ference by the traction current, reactance bonds, and the 
automatic signals are similar to those installed by the General 
Railway Signal Co. (the contractors for this work also) on 
the Electric Zone of the New York Central and Hudson River 
RR. (See Chapter XXL). 

Fig. 3550 shows a signal and in front of it a reactance 
bond fixed on two extended sleepers. The post on the right 
of the signal carries in the top case the line transformers, in 
the middle box the cast-iron resistance grids which limit the 
current-flow from the transformers above to the " Track- 
Circuits " when a train is standing at the transformer, and 
in the bottom box the track-relay and the terminal board. 
The illustration also shows the method of protecting the 
" live " rails. 

'g- 355 a - Automatic Signals, Philadelphia and Western Railroad. 




Switch Locks. 

. 356 illustrates a lock that is used in America for 
electrically controlling outlying sidings from signal-boxes. 

The rod a is coupled to the switches and passes through 
a box in which there is a latch b falling into a recess in the 
rod. When the points may be used an electric, current is sent 
by line wire from the signal-box which energises the magnet 
c, attracts the armature d, and raises the latch b, thus allow- 
ing the rod to be moved, and directly the rod is moved the 
recess is taken away, and the latch cannot return to its 
normal position until the rod is put back. 

A small lever e, at the end of one arm of which there is a 
roller / which engages in the recess, so that when the rod is 
moved it is forced out. The other end of e is fitted with 
contacts g l , g-, and these are drawn out and an indication 
thereby sent to the signal-box when the points have been 
used and when they are restored. 

been lowered by an approaching train the magnet c remains 
de-energised and the latch cannot be lifted. Fig. 357 illus- 
trates this. 

The releasing lever 4 normally stands "over" in the 
locking-frame, and has to be put back in order to release 
the points. To release this lever the next automatic stop 
signal in the rear (5) has to be at danger in order that spring 
5 may be closed. The lifting of the latch coupled to the 
clasp handle will open contacts 3, 4, and join 3, 5. The 
opening of 3, 4 breaks the circuit for relay N, and its arma- 
ture falls away, opening contact i and closing 2. The open- 
ing of contact i breaks the circuit of track battery T 1 , which 
in turn de-energises track battery B 1 , and its contact 6 opens 
and holds signal 7 at danger. The closing of contact 2 on 

Fig. 356. Hall Lock for Outlying Switches. 

In the boxes on the North Eastern R., that are only 
opened when any shunting has to be done, such a lock is 
fixed on the releasing levers, and the current from the battery 
to the automatic signals passes through the coils of the 
magnet c. The latch b lies in a short rod an interlocking 
tappet coupled to the lever, and when the clasp handle is 
grasped a circuit is set up which, if the two stop signals in 
the rear be at danger, takes the latch out of the short rod. 
At the same time the battery is cut out from the signals so 
that they cannot be lowered. If, however, the signals have 

relay N by the lifting of the lever latch completes the cir- 
cuit, and the lock f being withdrawn out of the tappet on 
the lever (similar to latch c in fig. 35(1) the lever can be 

It will thus be seen that any train that is shunting is 
protected by two stop signals and their corresponding distant 

' Track-Circuits " for Siding Connections and Crossovers. 
It has already been made clear that anything standing on 
the main line short-circuits the current for the signals so that 
they cannot be lowered, but provision has also to be made 
against any vehicle being left on a siding connection between 
the points in the main line and the safety points in the 

354- Diagram showing Electric Control of Hall Swilch Locks. 



Indicate Insulated Joints 

Fig. 358. Diagram showing Method 

siding, or between the two pairs of switches of a cross-over 

The diagram in fig-. 358 explains how this is done. The 
thick lines represent the "Track-Circuit" and the small 
squares are insulated joints. Consequently, the siding con- 
nection and cross-over are protected equally with the main 

Insulating Point Rodding, &>c. 

It is, of course, necessary to insulate point rodding, 
switch-rods, etc., so that they will not short-circuit. 

F'g s - 359 to 3 f ' 2 illustrate methods employed on some 
of the American RR. for insulating point rodding. 

of Insulating and Protecting Switch Points. 

r i'-Jf-2'--^t- 3" if -a" ,). 8" -2 $--- : -- 


*'i" 2 'i|"l r,^f.i.f 

\ e : ,| 

= 350- Insulating Point Rodding. 

? 359 shows that adopted by the Baltimore and Ohio 
RR. The switch rod is divided in the middle with an open- 
ing of Jin. and is provided with two plates. Two pieces of 
fibre of the same length are employed, which are laid between 
the plates and the switch rod. Round the four bolts fibre 
washers are provided. 

Fig. 360. Insulating Point Rodding. 

Fig. 360 shows the arrangement adopted on the New 
York Central RR. It is similar to that used on the B. and O. 
RR - (fig- 359), except that it has 6 bolts. 

Fig. 361 shows the arrangement adopted by the Lehigh 
Valley RR. The switch-rod is divided as usual, the two ends 

Fig. 361. Insulating Point Rodding. 

*4' 'Spring Coffer 


Fig. 362. Methods of Insulating Rodding and Joints. 
being flanged out as shown, with two thicknesses of fibre 
between. Fibre is also laid between the rod flanges and the 
washers under the bolt heads. Fibre bushes are also pro- 
vided for the two bolts. 


Where points are controlled by automatic signals, indica- 
tors are provided to show to the trainmen, whose train has 
to leave a siding, whether any train is approaching from the 
rear which they might delay, and whether the section ahead 
is clear. 

References are made to these in the standard rules of 
America as to the working of automatic signals (sec 

The indicators take various forms. Some are red discs 
which fall before a white opening when a train is in the 
section. Others have a miniature arm which is down when 
the line is clear and up when it is occupied. Other railroads 
have an electric bell (instead of, or in addition to, an indi- 
cator), which rings when the line is occupied. 

On the North Eastern R. these indicators take the form 
of miniature arms, and show when a train is in two sections 
in the rear and in one section in advance. 

For a crossover road indicators are required for both 


For making a contact when a lever is pulled or a signal 
lowered a commutator similar to that shown in fig. 363 may 
be provided. 

This is the Hall Co. 's commutator. The lever or signal 



- 3 6 3- Hall Commutator, 
arm is coupled to a crank on a shaft a. Secured 
to the shaft is a short arm b with a roller c which 
works in a spring d at the end of which is the con- 
tact blade e. The electrical connections are coupled 
to the binding screws / / which are insulated from 
each other, but the circuit is completed when the 
shaft a is turned as the blade e comes in contact 
with the jaw g. 

Switch Instruments. 

A switch instrument is provided to all points on and 
leading to lines protected by automatic signals, which 
guarantees that the points are properly closed, so that if any 
switch has not responded to the lever and is left open or 
partly open, or should any points have been run through and 
damaged, the signals remain at or go to danger. In 
mechanical signalling only the position of facing points are 
detected, and trailing points are in no way tested. 

- 3 6 4- Hall Improved Switch Instrument. 

Union Switch nnd Signal Co. 's 
Switch Controllers. 





''"iff- 366- Switch Controller. 

'ST- 36". Switch Controller. 



Fig. 368. Switch Controller. 


Fig. 364 illustrates a Switch Instrument made by the 
Hall Co. 

Attached to the binding screws are electrical connections 
in circuit with the signals to be controlled, and also the 
springs a 1 a- b l b- c l c- d l d- resting on rubbing contacts, two 
of which are continuous when normal and two insulated. 
But when over the former become insulated and the latter 
continuous. For instance a 1 a- rest respectively upon a 3 a 4 
which are separated, also b l b- in the same way on b 3 b*. 
But c 1 c- rests on c 3 , which is complete when normal, and in 
the same way d l d- are on d 3 , which is also complete. 

The switches are coupled to the crank e which turns the 
shaft / on which there is a cast iron segment g, on the upper 
surface of which is a wood fibre insulation covered by a 
German silver contact brush h. As the crank is moved by 
the switches going over, the contact brush h travels, and 
therefore the springs make fresh contacts, so that a 1 a- are 
joined together owing to a 3 a 4 forming one complete undivided 
contact, also b 1 b- in the same way. Contacts c 3 d 3 become' 
divided and insulated, so that c 1 c- and d 1 d- have their re- 
spective circuits broken. 


The Union Switch and Signal Co. recently introduced 
a new vertical rotary switch circuit Controller. A draw- 
ing of this is given in fig. 365. The rod a is coupled to 
the points and to the operating arm b. In the switch box is 
a horizontal segmental cam c, attached to the vertical pin 
i, and therefore actuated by the operating arm b. Across 
the switch box and above the cam is a pivot pin d on which 
turns the rocker c, having on one side a roller / resting on 
the cam c. At the ether side is a spring g in tension. Above 
all are four pairs ol contact springs h in the circuit with the 

signals affected. When the points are opened the operating 
arm b turns the cam c, which raises the roller /, and therefore 
depresses the other end of the rocker e and closes the con- 
tact springs and the signals are thrown to, or kept at, danger. 

One advantage of this form of controller is that owing to 
the segmental shape of the cam there is no interference of 
the contact switches by the springing of the point switches 
as trains run through them. Its height is only 7ins. from 
the/base of the rail. 

The object of using these controllers and the way they 
perform their functions will be better understood from the 
diagrams, figs. 366, 367 and 368. 

Fig. 366 shows a siding having a connection with the 
main line. "Track-Circuit" is provided for the lines 
shown thick, the track battery being at B. The circuit flows 
through the switch box a and when the points are moved |th 
of an inch the contacts b 1 b- b 3 b 4 are closed, thus short cir- 
cuiting the relay R on the signal and causing it to go to, or 
remain at, danger. The " Track-Circuit " ends at the 
fouling point of the siding, which should be provided with 
safety points or derails. 

The switches are provided with wedges c 1 c- on the nor- 
mally open side and which lift them clear of the tie plates (or 
chairs if used) and so render it unnecessary to electrically 
cut out the switch by insulated joints in the main line. 

I' ijJr- 3''7 shows a crossover between the two main lines, 
and here there are two switch-boxes by which the correct 
position of the points in each line is guaranteed. 

Fig. 368 shows a siding connection as worked from a 
ground frame <i with a switch and lock movement b 1 b 2 on to 
which the circuit controllers c 1 c" are attached. 


There are many patterns of Relays. Not only are they 
used for various purposes, but all signal manufacturers have 
their own types. For instance, there are neutral and 
polarised relays for " Track-Circuits '.' on steam-worked 
roads and polarised relays for alternating current and for 
direct current on electrically worked railways. Relays are 
also used for working indicators and for numerous other 

1-lg. 369 represents the \Vestinghouse neutral track relay 
for steam roads, and fig. 370 the \Yestinghouse polarised 
track relay for electric roads. 

It is well-known that relays are delicate pieces of 
mechanism, and amongst other safeguards against inter- 



Fig. 369 
Neutral Relay 

Steam Railways. 

Fig. 370 
Polarised Relay 

Electric Railways. 

fcrcncc with their adjustment they are generally enclosed in 
airtight, sealed cases. 

The Sykes relay, illustrated by fig. 371, j s enclosed in 
an oil bath, in a glass vessel G, which has a wooden cover 
H and between which and the lip of the vessel is a pad h 
of felt. A wooden base K is provided and the top and 
bottom are joined together by the bolts J, a pad k of felt 
being placed under the vessel between it and the base K. 

The switch A, with its limbs a a, is suspended by means 
of the link c from one arm C of a lever pivoted at d, to the 
longer end of which is coupled by means of the short rod b 
the movable core or armature F of a vertically disposed 
solenoid E. When the latter is energised the armature is 
raised, and consequently the limbs a a enter the mercury 
cup and complete the circuit. 

The wires I I and b b enter at the terminals L M and the 
apparatus is carried on a base g. 

Insulated Joints and Replaccrs. 
These are described in Chapter V. 

Signals for Level Crossings of an Electric (Over-head) Line 
and a Main Line Rail-way. 

The Burton and Ashby light railway crosses the main 
line of the Midland R. near Ashby on the level, and the way 
in which the signalling was carried out by Messrs. Brecknell, 
Munro and Rogers is illustrated by figs. 372 and 373. 

Hxcd on the trolley pole, nearest the crossing on each 
side, is a box a, a signal arm b and a lamp c with red (upper) 
and white (lower) lights. The railway company have a 
signal box near in which a lever d in the locking frame is 
reserved for operating the signals from the signal-box. The 
lever here shown does not correctly represent the actual lever 
used, so it suffices to say a lever d is provided. 

Two solenoids e- e z are fixed on the pole to which arc- 
attached rods f- f, the former for lowering the signal arm 
and the latter for raising it. Situate between the rods f f 
is a rotary switch g from which the electrical connections 
are run to the solenoids and to the signal lamps, and thence 
to pilot lights in the signal-box. 

x fe 

Fig- 3/1- Sykes Relay. 

. operating 

Signa I A rm 

_^J lt_ 



. /--v fl - 


*" " To Signal on other side of 

r 1 

' ' 

1 o 

: o ! 

: <> ; i 9 

' 9 

Signal Lamps 
on Pole 
Pilot Lamps 

DtaGram of Connections 


/n Cnbin 

j - L cad to Red ligh t 
f Lead to white lia,M 

implex Tube to Lamp 

v - To sere ,n Bo* 
Box bonded to Pole 

if ml Arrangement Wiring Diagram of Solenoid But 

Fig. 372. General Arrangement of Semaphore, Lamp and Box, and 
Connections, for Level Crossing Signalling. 




- 373- Details of Rotary Switch, Trip Gear and Locking Gear. 

The rotary switch, which is illustrated in detail by fig. 
373, is fixed on a shaft, on the end of which is a trip with 
three arms, each of which has a weight at the end g 2 , a 
heavy weight and g 3 g 4 lighter ones. 

When the lever in the signal-box is pulled over from 
position d- to position d 3 current flows to contact g 5 on the 
rotary switch and thence through g 6 to the solenoid e 2 , 
energising it so that the rod f is lifted and the signal arm 
lowered. The tail of the arm and rod f on rising raises the 
weight g 3 , and consequently the heavier weight g* which, 
as soon as it gets over the centre, falls to the other side, 

when the other smaller weight g* engages with the rod f 
and presses it down and so assists in lowering the signal. 
This movement has turned the rotary switch g, and contact 
is now made with g 7 to the signal lamp and the pilot light 
in the signal-box. 

Similar connections are made with the signal on the 
opposite side of the line. 

\\ hen the lever d is put back to position d 2 solenoid e 3 is 
energised, e- is de-energised, rod f is raised, the signal goes 
to danger, and the rotary switch reversed. 



THE number of arrangements for repeating fixed signals 
in the cab of a locomotive that have been patented is legion. 
These have generally, if not always, been designed for sig- 
nalling during foggy weather. 

That there is need for some such signal in foggy weather 
is admitted, but there have been difficulties in the way of 
the successful application of such a system, not the least of 
which is the necessity for practically its universal adoption 
owing to the extensive exercise of running powers by foreign 
companies. A good cab signal would be useful in all 
weathers and at all times. 

This class of signal is divided into purely mechanical 
ideas, like Raven's on the North Eastern R., and those which 
bring electricity or some other power to their aid. It seems 
to the Author that the latter class comes within the scope of 
this work. 

Boult's System. 

One of the oldest, and best known, systems is that in- 
vented by Mr. W. S. Boult. It has been tried on both the 
Gt. Northern R. and the Gt. Central R., but further pro- 
gress has not been made on account of, the Author believes, 
the cost which the adoption of the system would entail. The 
apparatus employed is very complete and every condition of 
working seems to be provided for. 

An essential part is the attachment to each engine of the 
necessary electrical equipment, bells and indicators. Mr. 
Boult recognised that there were objections to signals on 
the engine being given by means of contacts with conductors 
fixed on the line. The blow given at high speed, the wear 
and perhaps bending of the parts would all lead to want of 
alignment, so that in time contact would be missed. Dirt, 
snow and frost would also be likely to affect the contact. 
The inventor, therefore, turned his attention to magnetism 
as a means whereby signals could be transmitted, and he 
designed a system consisting of magnets fixed on the per- 
manent way so arranged as to establish a long magnetic field 
of transverse polarity, and arranged on the engine an iron 
armature containing, in a divided gap or space, a small 
pivoted needle on which magnetism collected by the arma- 
ture is condensed or focussed so as to strongly act upon it 
and throw it over to one side. 

Pig. 275 is a diagram of the magnets. Under the chairs, 

and in magnetic connection therewith, are plates of iron A 3 
and upon these plates and also upon the sleepers rest blocks 
of wood B which support a piece of timber C placed edgeways 
upon them and furnished with holes (> in which are placed 
permanent magnets D or electro-magnets D 1 . 

The armatures carried by the engine are F and these serve 
to collect and focus the lines of force from the magnets upon 
a deflectable needle, situated, as already said, in a gap in 
the armature. 

Each of the armatures comprises three collectors F 1 cai- 

jrr; \', I I ',/ \V V M\ ! I if /'/ 



^s ^ 




Fig. 375. Boult's Cab Signal. 

ried upon the engine in such a position that they pass over 
the permanent magnets, the rails and the electro-magnets 
respectively. In the spaces between the collectors are bridge 
pieces F 2 and boxes F 3 which contain four needles, two ol 
which are actuated by the danger or " on " position and two 
are for the " off " position. Either of the two pairs of needles 
are sufficient to give the signals they are intended for. In 
fig. 375(1 is given a view of the indicator on the engine. 

Returning to fig. 375 it will be noticed that the permanent 
magnets are fixed fan-shape in the timber, with their lower 


20 1 

ends standing in groups on iron plates on the sleepers. The 
iron plates extend under the chairs so that the magnetic 
circuit is through magnet, outer pole-plate of armature, 
bridge piece and ne'edles, inner (central) pole plate, railway 
metal, chair and iron plate to magnet. 

lamps) are placed in the cab of the engine. These lamps 
are actuated by the armatures of electro-magnets energised 
according to the state of the "Track-Circuit." The most 
serious question is, whether or not the Miller cab signal is 
intended to be auxiliary to, or independent of, fixed signals? 

Fig. 3750. Boult's Miniature Cab Signal. 

The electro-magnets are fixed further along the line and 
act in the same manner. These are energised according to 
the position of the signal they indicate for. 

The method of working is : at some distance from each 
stop signal is fixed a set of permanent magnets which throw 
both arms of the indicator to "on." Immediately in ad- 
vance are the electro-magnets which, if energised by the 
signals being " off," cause the miniature signals to be 
lowered, to correspond. Should it be a distant that is being 
passed and the distant be "off" the lower arm will drop 
to indicate this, but if the signal is " on," then the lower 
arm remains " on." The same happens with the upper arm 
when a stop signal is approached, and one advantage of the 
Boult system is that an indication to a driver standing at a 
stop signal is given when the line is cleared and the signal 
lowered for him to proceed. The four upper circles on the 
left of the indicator are to show to a driver approaching a 
junction what road is set for him. Four sets of figures are 
used MJ, i, 2, 5, the first always standing for the main 
line, whilst i stands for a branch line at an ordinary double 
junction. \Yhere there is more than two lines to be travelled 
upon i and 2 together indicate road 3, road 4 would be Ma 
and 2 and so on. These indicators are actuated by a second 
set of electro-magnets. 

Boult's system is not only applicable to working during 
fog but in all states of weather, and amongst its advantages 
is the absence of all movable parts on the permanent way, 
no additional weight on the signal, no complications of slotted 
signals, no obstruction to be struck and signals given at any 
rate of speed. 

Miller's Cab Signal. 

This is an American system which was for some years in 
work on the Chicago and Eastern Illinois RR., and in the 
Park Avenue Tunnel, New York City. In the former case no 
outdoor signals previously existed, and the Miller signals 
took the place of them. At Park Avenue the Miller signals 
were in addition to the existing fixed signals. 

An essential feature of the Miller System is the adoption 
of the "Track-Circuit." Electrical signals (red and white 

Fig. 376. Insulation of Engine, Miller Cab Signal. 

In other words, is the cab signal to be used in place of 
fixed signals, as on the Chicago and Eastern Illinois, or in 
addition to them, as an aid to the driver? In the latter event 
the indications would be very useful to a driver, and in 
foggy weather the greatest possible boon. It would also be 
an advantage to a driver to indicate to him the state of the 
signal he has just passed, in case he omitted to notice it or 
had forgotten its condition. Or the signal may have been 
passed at "danger" and subsequently have been cleared, 
in which case the indicator on the engine would show "clear" 1 
also. But here in this very advantage apparently a great 
one lies a difficulty. Is a driver, having passed a distant 
signal in the "on " position, authorised to subsequently 
ignore that signal should his indicator go to "clear'"? 
Again, if a distant signal be " on " and the indicator shows 
" line-clear," what course should a driver pursue? It cer- 
tainly seems wrong to tell a man to ignore his fixed signal, 
and no officer would do that ; and yet should the signal be 
' off " and the indicator show danger, that is the course 
he would have to take. Of course it is easy to meet the 
difficulty by instructing the men that when contrarv signals 
are shown, to prepare to stop. 

Fixed signals serve other purposes than giving indica- 




ig. 377- M' llel " Cab Signal System. 

Block B-C 

Black Station A 

tions to drivers. They notify the guards of trains, they 
give warning to men working on the line of the approach 
of trains, and in case of a breakdown it is easy to see if any 
obstruction be protected, because the fixed signals are 
visible. Such a system as the Miller cab signal gives no 
guarantee that the obstruction is properly protected. 

This system was introduced into England in February, 
1903, when a trial installation was laid down on the west 
side of Woodhead Tunnel on the Great Central R. This 
trial was so far satisfactory that the Board of Trade sanc- 
tioned its being tried on the up line through Woodhead 
Tunnel on condition that absolute block working was used 
for passenger trains and that the Miller system was only 
for protecting goods and mineral trains. 

The installation in Woodhead Tunnel was not, however, 
laid down, and now, as stated in Chapter XIII., low-pres- 
sure automatic signals have been fixed there instead. 

After the trial at Woodhead a fixed signal has been re- 
garded by the Miller Co. as part of their equipment. 

Before the addition of a fixed signal the system consisted 
of automatic signals on the "Track-Circuit " principle being 
given on the engine by means of red and white electric light 
bulbs, the current for which is generated by a small dynamo 
carried on the engine. 

The line is divided into the usual block sections with 

' Track-Circuits " and at the entrance to each block section 

there is a short length of insulated track, the first two insulated 

joints of which are opposite each other, but the other two 

are staggered. 

The engine is insulated from the tender as shown in fig. 
376. The draw-bar is insulated at A, the buffer plates at 
B B, the water pipe connection is insulated at C by putting 
a gasket of hard fibre between the flanges of the pipe con- 
nection, the foot-plate is insulated at D and the safety chains 
E are insulated by substituting a special insulated link for 
one of the regular links. 

The insulated wire leading from the tender to the signal 
bulbs is at F whilst K is the insulated joint and L the signal 

Should the engine not have a tender then the trailing 
bogie is insulated instead. 

The remainder of the electrical connections are shown in 
fi g- 377> where a block section from A to B is shown. The 

Line Wire A~B Block Station B 

instruments at A and B are precisely similar. A magnet d 
has its coil connected across the rails at the entrance end 
of A B, and when that block is clear the current from the 
' Track-Circuit " battery c energises the magnet. The mag- 
net has two armature contacts : one of these, /, communi- 
cates with the preceding block through line wire and rail, 
by way of m, h and 11 ; the other contact, e, communicates 
with block section A B by way of line wire and rail, through 
the coils of a second magnet g. This latter has an armature 
h which acts as a reversing switch for battery j, and con- 
nects this battery with the points H and i of the rail, on 
opposite sides of an insulated joint. When, armature h is 
attracted, as it is represented in fig. 377, then the current 
from y tends to flow in the rail from point m to point n; 
when h is dropped, the current tends to flow from point n 
to point m. 

The operation of the relay instrument and its electric cir- 
cuits may now be readily followed out. When the track A-B 
is clear, then magnet d is energised by track battery c and 
attracts its armature. Similarly, when block B-C is clear, 
magnet d 1 at B attracts its armature. There is then an 
electric circuit, as shown, from battery j 1 at B, through /; l 
and wire m 1 , through lever f 1 back to A, energising magnet 
g, and back through the rail. Armature h is attracted when 
magnet g is energised, and an electric pressure is set up 
from point m to point n, which affects the signals on the en- 
gine as explained below. When either A-B or B-C is occu- 
pied, then the circuit through the coils of is broken, at e 
or at f l respectively, and armature h drops, so that battery ;' 
tends to send its current from n to m through the rail. This 
pressure between n and m gives an opposite signal on the 
locomotive equipment when the engine crosses the joint be- 
tween n and m to that when the pressure is from in to n as 
explained below. 

One axle each of engine and tender, h and 7, are insulated 
from each other, and are connected through the coils of two 
magnets, p and q. The armature which plays between the 
poles of these magnets is adapted to make contact at .v or at t, 
completing the circuit of a battery u (generated by a small 
dynamo on the engine) through the white lamp or the red 
lamp, respectively. When the engine crosses the insulating 
joint above referred to, the pressure between n and m causes 
a current to flow through the coils of magnets p and q, and 



the direction of the current between in and n determines 
whether the armature is attracted up or down, to light the 
white or the red lamp. \Yhen the current flows from n to m 
then the armature is attracted downward, lighting the red 
lamp and indicating that either block A-B or block B-C is 
occupied. When both blocks are clear the battery / sends 
current from m to n by way of axle / to fe, so the armature 
is attracted upward and the white lamp lights up. 

It will be noticed that a small battery r is arranged on 
tin engine for a local circuit through the coil of magnet q, 
in such a way that no polarised armature is required, while 
at the same time the battery r acts to hold the armature in 
place, either up or down, while passing through the block. 
When an insulated joint is reached the circuit r through 
magnet p, by way of the rail back to /, is broken, and mag- 
net q being still energised, the armature drops and lights 
the red lamp. Thus the engine always gets a red signal the 
moment it passes an insulating joint, unless at the same 
moment the " clear " impulse from the track relay operates 
to keep the white lamp alight. In order to avoid this action 
while crossing the insulating rail joint at the rear end of 
block A-B, this joint in the two rails is staggered longitudin- 
ally a distance of half a rail-length, or at least a distance 
greater than the interval between axles k and /. 

Sheehy's System. 

A most ingenious system has been invented by Mr. Robert 
J. Sheehy, of Xew York, but like the great majority of such 
systems, it does not appear to have got beyond the working 
model stage. 

Signals are given on the engine showing when an ob- 
struction exists on the line ahead of the driver and whether 
the obstruction is due to another train or misplaced points, 
broken rail or open drawbridge. A time-recording device 
automatically prints a record of the hour, minute and second 
a signal is received and of the time intimation was given 
that the obstruction was removed, hence the time lost is 
recorded and a history of the run given automatically. When 
the line is clear no signal is given, but it is shown when an 
obstruction is approached, steam being at the same time shut 
off and the brake partly applied. The signalling system is 
also stated to be equally applicable to the safe working of 
single as of double lines, whilst it is claimed to deal effectively 
with the difficulties previously pointed out herein as associated 
with automatic signals on electric roads. 

1 he above is only a brief summary of the numerous ends 
Mr. Sheehy has aimed at, and on the model achieved. 

Kinsman's Automatic Train Control. 

In the grounds of the exhibition held in connection with 
the International Railway Congress at Washington, 1905, 
the Kinsman Automatic Stop for steam-worked trains %vas 

On the line a contact maker is fixed which works with 
the signal so that when a signal is at danger the contact 
maker is in position to make contact with a corresponding 
contact on the engine. 

In fi g- 37 8 tn e regulator b is open. When a signal is 

passed at danger the magnet c is energised and armature d 
attracted, so that the lever c falls to the right and this turns 
the chronometer valve / so that air passes through pipe g 
from the train pipe and forces the clutches h h oft the slightly 
bevelled side of depression / and frees the follower fe, which 
can be driven to the left independently of the regulator con- 
nection, so that it is forced to the left and closes the regulator 
b and shuts off steam. The air from the train pipe then 
escapes through port /, so that the brakes arc applied. 

''"' 37^' Kins. nan's Automatic Control. 

\\ hen, now, the driver reverses his regulator handle a 
the valve m is opened and air admitted to the cylinder n 
so that the piston o, which was forced in when the chrono- 
meter valve / was turned, is driven out, and the chronometer 
valve automatically closed and pressure in the train pipe is 
made normal. The lever e is also restored. 

But .it is in connection with its equipment on the cars of 
the Xew York Interborough R. (the Subway) that the 
Kinsman automatic stop is best known. The automatic stop 
there used was described in connection with the Subway 
signals in Chapter XIV. These are electrically worked 
trains and a similar magnet and armature releases a lever 
which opens a chronometer valve which releases the air from 
the train pipes so that the brakes are automatically applied. 

La/fas' s Automatic Train Control. 

One of the first successful automatic controls tried in this 
country was the Laffas. 

This consisted of an obstruction in the four-foot that was 
parallel with the rails and was turned and raised by a 
weighted connection from the distant signal when in the 
" on " position. This struck a trigger on the engine and, 
opening a valve, automatically applied the brake. 

The Laffas system had a successful exhibition on the 
Barry R., but it was never adopted by any company. Pos- 
sibly one reason for this lies in the fact that an automatic 
application of the brake at a distant signal is not always 
desirable. For instance, a driver may be approaching a 
terminal station with his train under proper control, but were 
his brake applied at each distant signal he passed, and he 
may have to pass six or eight, it would mean no end of 
delay, as not only would the brake be applied, but it would 
have to be released. Another objection is where distant 
signals arc on a rising gradient. To pull up a goods train 
there might lead to its being stalled. 



Jefcoate's Cab Signal. 

Of the cab signals with:: ut automatic controls, one of the 
simplest is that invented by Mr. H. J. Jefcoate, of Crewe, 
and which is illustrated by fig. 379. 

The mechanism is as follows : 

At the distant signal a contact bar z is laid, and at the 
home signal two contact bars 4 and 6 are laid. On the 
engine three plungers a 1 a 2 a 3 are fixed, and these complete 
electrical circuits and light lamps, raise small semaphore 
arms on the foot-plate in front of the driver, and sound bells 
of different tones. 

green light, the " home of! " semaphore on the foot-plate 
and continuous ringing of the bell until stopped by the driver 
switching off the current. Fig. 379 also shows a detail of 
the plungers which are especially designed to avoid shocks 
when the contact is made. The plungers are mounted in 
ball bearings and are arranged to rise only ^in., and ample 
allowance is made for oscillation of the engine. 

The giving of an intimation to the driver, after he has 
been pulled up, that the signal has been lowered is a step 
in advance of anything yet attempted. 

Raymond Phillips' Automatic Train Control. 

The principal objection to " obstruction " signals is that 
the breakage of the " obstruction " or tripper on the line 
would lead to the warning not being given, whilst the break- 
age, unknown to the engine men, of the corresponding part 
on the engine would probably lead to disaster. 

The idea of automatically applying the brake is a very 
good one, but it must be arranged with care, and if a train 
has to be pulled up it should be done at the home-signal, 
but that may be too near and possibly too late, and, there- 
fore, a slight application of the brake should be made at the 
distant signal and a full application at the home. 

In the Raymond Phillips system, which was tried at New- 
castle Junction, N.S.R., most, if not all, the objections above 
noticed have been met. 

That portion of the apparatus wherein the greatest risk 
lies is the lever on the engine that strikes the tripper in the 
track. In the Raymond Phillips method the lever is so con- 
structed that in the event of it being broken the brake would 
be applied and the driver unable to release himself without 
first replacing the lever, a question though of only a few 

Fig. 380 illustrates the mechanism fixed in the " four- 
foot " near the distant signal. In the lower part of each 
case is a slide a coupled at b to the distant-signal wire, or in 

n n 

!"' 379- Jefcoate 

Supposing the distant 3 to be " on," the contact bar con- 
nected to the signal wire would be raised, and on the plunger 
a 1 striking the bar, a circuit is closed which gives the driver 
a red light, shows the " distant on " semaphore, and sounds 
a bell, which keeps ringing until it is stopped by the driver, 
who would take steps to get his train under control so as 
to stop at the home signal 5, which we will assume is " on " 
when he reaches it. The contact bar 4 being raised, plunger 
a" makes contact with it and the driver gets another red 
light, raises the " home on " semaphore, and the bell rings 
only so long as he is running over the bar, but when he 
comes to the end of the bar the bell stops and the train 
should also be stopped. When the signal is taken " off " t he- 
bar 4 is lowered and the bar 6 raised, the latter thereby 
making contact with the plunger a 3 and giving the driver a 

's Cab Signal. 

the event of the distant being a controlled signal to the 
signal rod. The other end c of the slide is attached to a 
balance weight to ensure the slide returning to normal. The' 
trippers d d have a weight c to keep them upright, and they 
can be turned in either direction as struck by an engine. The 
levers are in duplicate, as one might be broken or not make 
proper contact. When the slide a is drawn to the left, owing 
to the signal being lowered, the stop / on the slide comes 
against the weight c, and so turns the lever d to the right and 
clear of the lever on the engine. If, however, the signal be 
" on " the tripper d is struck by the lever on the engine. 

In the signal-box, near the lever in the locking frame that 
works the signal, is an electrical repeater with visual indica- 
tions as to the position of the. obstruction. To work this 
there are electrical contacts on the tripper d, and thereby the 



signalman is told if the lever be in proper position when the 
signal is "on " and whether it has responded to the clearing 
of the signal when the latter is lowered. Further, and this 
is a very important point, a bell is rung whenever the re- 
peater is changed from " on " to " off," whether this be 
done by the action of the signalman or by the temporary 
deflection of the lever, due to its being struck by an ap- 
proaching train passing the signal in the " on " position. 
Furthermore, this bell continues ringing until the signalman 
stops it. The man is thus impressed with the knowledge of 
the approach of a train, a fact that is of great importance 
at night time and in foggy weather, and such an intimation 
would be extremely useful on those rare occasions when a 
train enters a block section without being signalled on the 
block instruments. 

Fig. 380. Raymond Phillips' Tripper. 

A similar tripper with electrical repeater, etc., is fixed at 
the home signal but in a different line to the tripper at the 
distant signal so as to strike a second lever on the engine. 
The apparatus may also be fixed at any other stop signal, 
and here it may be said that where there are junction signals 
only one tripper c:ise need be fixed. 

Fig. 381 illustrates the gear on the engine. Suspended 
from the framework is a case in which are two levers similar 
to b, one end of each of which projects from under the case 
and comes in contact with the trippers at the signals, one 
lever being applicable to the distant signal and only applying 
the continuous brake sufficiently to check the train and the 
other lever being applicable to a stop signal and applying 
the brake sufficiently to pull the train up. In the cab of the 
engine are two gauges, r 1 c-, containing indicators repre- 
senting distant and stop signals, c 1 being the dis- 
tant and r 2 being the stop, or both indicators may 
be in one case with upper (stop) and lower (dis- 
tant) arms. When running these arms are down 01 
" clear " as the vacuum reservoir is normally in connection 
with the gauges through the valves similar to d which is the 
distant valve. Should a distant signal be passed in the "on" 
position the lower part of the distant lever b would be turned 
to the right as shown in dotted lines and the upper part to 
the left. This would cause the weighted end of another 
lever, e 1 , to drop and fall behind the upper part and so hold 
it. This would allow the piston valve d to fall, so cutting 
off the vacuum from pipe / and opening pipe g to the atmos- 
phere, which would cause the miniature arm c 1 to rise. The 
other end of lever e l is coupled to a piston working in cylinder 
h l which would then be raised and allow air to enter the 
train pipe through the syren and pipe ;'. This causes the 
syren to sound and the brakes to be partially applied. In 
order to release the brake and stop the noise of the syren 
the three-wav cock k has to be turned and this connects the 

reservoir with the lower side of cylinder h l through I, m so 
that the piston falls and the lever e 1 returns to normal and 
lifts its weighted end from holding lever b off, which would 
then resume the perpendicular. As the lever e 1 resumed its 
normal position it would restore piston valve d and re-open 
pipe / and close g and the distant arm would fall again. 
Matters would then be normal, but to do this the driver has 
had to turn the three-way cock, otherwise the syren would 
continue sounding and the brake applied. 

For the stop signal there are also levers, and a cylinder 
and similar applications of the brake and sounding of the 
syren. The lever corresponding to e 1 is e 1 , and the cylinder 
is h". There is this difference though the stop signal lever 
e- controls (or slots) distant lever e 1 , so that both miniature 
arms will go to " on " when the stop arm is raised. This 

TbSyren < 

IbTrffrn Pipe 

Fig. 381. Raymond Phillips' Train Control. 

is to make this form of signalling to agree with standard 
outdoor signalling, and is obtained by using a T lever n 
like the mid-lever of the slot of a signal, and suspending 
the distant piston valve d over the T lever, so that when the 
lever e 2 falls it carries with it lever n, so allow- 
ing valve d to fall, but when lever e 1 falls it 
carries with it lever n, but not lever e-. It follows 
also that a double area is open to the atmosphere 
when a stop signal is passed at danger, which causes a 
quicker application of the brake and which cannot be readilv 
overcome bv the driver. 



The lever of the three-way cock is normally downwards, 
and it is heavily weighted to secure its restoration to that 
position when turned, so that the bottoms of the cylinders 
h l 7i 2 are freely exposed to the atmosphere to ensure that 
their pistons rise freely when the levers e l e 2 are operated. 
The pistons are suspended, so that in the event of anything 
failing they will fall and apply the brake. 

The striking levers, whilst mounted on the same axle, 
are on separate bushes. Each is made in two parts, like a 
pair of scissors, and the levers are so constructed that if they 
broke or if the end were knocked off, the two parts would 
be forced open by the spring o, and so the weighted end 
of the lever would fall into the opening and apply the brake. 
Or if the lever were swept away from any cause or any other 
failure occurred, the piston d would fall and apply the brake. 
Herein lies the security against a driver running unconscious 
of the fact that his apparatus was defective. 

Briefly then the system may be stated to provide an 
audible and visual signal when a driver passes a distant or a 
stop signal in the " on " position and an indication is given 
to him whether the signal is a stop or a distant. That the 
train is put automatically under control according as to 
whether it is a " stop " or a " distant." That an intimation 
is given to the signalman, also by audible and visual signals, 
when a train passes a signal in the "on " position. Lastly 
safeguards are provided against failures of the mechanism. 

Western Syndicate System. 

Mr. R. J. Insell, the signal assistant, and Mr. C. M. 
Jacobs, the electrical assistant, to Mr. Blackall, the signal 
engineer of the Great Western R. have been the principal 
designers of an apparatus which, after a trial on the Henley 
double-line branch at all rates of speed, has been laid down 
on the Fairford branch, which is single throughout, and has 
eight signal-boxes in its length of 22 miles. 

K c^t 

Fig. 382. Western Syndicate Control. 

It is of the cab signal type and dispenses with distant 
signals entirely. It was brought into use in December, 
1906, and the results of such a radical departure should be 
watched with interest. 

Fig. 382 is a side elevation of the arrangement and fig. 
383 shows a cross section of the ramp on the permanent way 
and of the contact on the engine. The signals are given by 
a whistle indicating that the stop signals ahead are at danger 
and by a bell ringing to indicate that they are at "clear." 
Visual signals are also given on an indicator in the cab. 

Fixed on a timber base is a ramp r of T-iron and from 
4oft. to 6oft. in length, and the top of which is 4ins. above 
rail level. On the engine is a shoe I insulated from the 
engine and to which a switch s is connected. The bottom 
of the shoe is 2jins. above rail level, and therefore when the 
ramp is struck the shoe is raised i^ins. and this breaks the 
circuit between the local battery b and electro-magnet e so 
that the armature k falls. This opens valve v and sounds 
the whistle which continues until the driver shuts it by mov- 
ing lever Z 1 . 

The whistle does not, however, sound should the signals 
be " off " and this result is achieved in the following manner. 
The ramp is connected by a line wire to a switch s l in the 
signal box and thence to the battery b l . When the stop 
signals are lowered and the signalman wishes to indicate 
this fact to a driver he pulls over the lever that under usual 
conditions would lower the distant signal. This connects by 
the switch s 1 the battery b l to the ramp r. The shoe Z is 
connected on one side to the magnet e 1 which, in turn, is 

Fig. 383. Western Syndicate Control. 

connected to a polarised relay p, the armature of which, when 
attracted, will close a local circuit between the battery b- 
and the bell a. Consequently, therefore, when under the 
conditions just named, viz., the joining up of the battery 
b l to the ramp r by means of the switch s 1 , the shoe Z comes 
in contact with the ramp r, magnet e 1 is energised so at- 
tracting armature k and therefore keeping the whistle closed 
the relay p will be energised and a circuit completed be- 
tween the battery b- and the bell, so that the latter rings, and 
it continues ringing until stopped by the driver pressing in 
push button r. 

The normal condition of affairs, then, is that when the 
shoe strikes the ramp a " danger " signal is shown, and this 
provides for snow and ice on the contacts or failures in the 
electrical apparatus. 

For single lines where trains have to pass over a ramp 
when they leave a station or passing place, and which does 
not apply to them, special arrangements have been made. 
If a single line section, say, between A and B be assumed, 
and a train is leaving A for B, it would have to pass over 
the ramp on the B side of A, and this would give an un- 
necessary signal and possibly a misleading one. To blow the 
whistle would be wrong, and to sound the bell. But it has 
already been said that either of these events must happen. 
There is, however, a way of escape, which is to send a 
positive current to the ramp. This would energise magnet 
e 1 and attract armature fe, so that the whistle was not blown 



but the bell would not ring. This, then, has been done. It 
has required some scheming in order to guard against the 
signalman improperly making the ramp "dead," and it has 
been secured by the lever and circuit being controlled by the 
circuit whereby permission is received to take out a single 
line electrical staff or tablet. 

The switch D is provided to cut off the local battery 
when the engine is out of running condition. The switch 
opens when steam in the boiler falls to 20 Ibs. pressure. 

One of the several advantages of this system is that the 
ramp may be placed in any situation regardless of distance, 
curves, bridges, tunnels, etc. 

Bonneville and Smith's System. 

This system of electric bell and miniatnre cab signal is 
being experimented with on the Beckenham-Xorwood branch 
of the South Eastern and Chatham R. 

Raven's System. 

For some years the cab-signal invented by Mr. Vincent 
L. Raven has been in regular use on the North Eastern R. 
main line. It is referred to on p. 158. The system has two 
drawbacks, viz., it does not indicate "clear" signals nor does 
it differentiate between stop and distant signals. 

A new method, also invented by Mr. Raven, has now been 
laid down for a distance of 14 miles near Xewcastle and 20 
engines have been equipped with the cab-signal. In this 
method a series of ramps are laid in the "four-foot" between 
the distant and the stop signals and a ramp in the "four-foot" 
and one outside each rail at the distant signal. On the engine 

there is a rotary switch to make contact with the outside 
ramps and a steel brush to make contact with the ramps 
between the rails. An indicator, with bell, is provided on the 
footplate and there is also a visual signal as to how the road 
is "set" at a coming junction. The ramps are joined up 
electrically when the levers in the signal-box are over and 
when the engine passes over them the circuit is completed 
whereby a bell is rung and a visual signal given. Should the 
levers be normal the bell is rung and another signal is given 
indicating "danger." 

An advantage of there being more than one ramp is that 
the signals continue to be given right up to the box, so that if 
a "danger" intimation be given at the distant signal and be 
subsequently changed to "clear" the change is promptly 
given on the engine, and in the event of the signalman 
reversing his signals from "clear" to "danger" that change 
is also promptly given on the engine. 

Automatic Control on Electric Roads. 

In Chapter XIV. the automatic controls on the Boston 
Elevated, Xew York Subway, Philadelphia Subway, Metro- 
politan District, and the " Bakerloo," " Piccadilly " and 
" Hampstead " tubes and the Waterloo and City railway 
were described. 

It may be remarked that the Board of Trade agreed to 
converging District electric trains being accepted by the 
signalmen at Turnham Green and Mill Hill Park at the same 
time, as there was room between the home signals and the 
fouling point for the train to be automatically pulled up in case 
the motorman overran the stop signal. 

T- 3^4- The first Power Plant designed by Mr. Geo. Westinghouse, 
laid down at Bound Brook, Philadelphia and Reading R.R., 
now preserved at the Swissvale Works of the Union 
Switch and Signal Co. 



IN an ordinary signal-box the work of the signalman in 
shifting over points and lowering and replacing signals is 
compensated by intervals of rest between trains. But in 
large signal boxes, where there is a constant succession of 
train-movements the labour performed by a signalman even 
if he be only employed for eight hours is very heavy. 

The greater size of the stations and yards of the present 
day and the later and better practice to provide signals for 
all possible movements has led to a great increase in the 
number of levers that are required in locking-frames, and the 
consequent increase in the size of the signal-box. 

Land has simultaneously risen in price, so that whilst signal- 
boxes have increased in area, and make the difficulty of 
finding suitable space for them the greater, there has been a 
coincident rise in the price of such land which has, owing to 
its value, probably been laid out to the best advantage pos- 
sible by the engineer. 

The operation of points and signals by power at once 
touches these difficulties. The signalmen's labours are very 
considerably modified, and the levers controlling the move- 
ments are small and placed closer together, so that the size 
of the locking-frames is very materially reduced, and conse- 
quently the area covered by the signal-boxes. 

To Mr. George Westinghousc must be given the credit 
for being the first to successfully operate points by power, 
and the Italian firm of Bianchi-Servettaz were next. The 
first power signal plant was an all-air one designed by Mr. 
Westinghouse and fixed at Bound Brook on the Philadelphia 
and Reading RR. of New Jersey in 1884, whilst a Bianchi- 
Servettaz hydraulic plant was fixed at Abbiategrasso on the 
Mediterranean R. in 1886. 

This interesting, though not yet ancient relic, is now 
preserved at Swiss Vale, Pa., by the Union Switch and Signal 
Co., to whom the Author is indebted for the photograph from 
which fig. 384 on the previous page has been reproduced. 

High-pressure compressed air, low-pressure compressed 
air, water, electricity, air and water, air and electricity, and 
electricity combined with manual, have all been, and most 
of these agencies are now used for the operation of points 
and signals. 

The earlier Westinghouse "All-Air" plants were soon 
altered, so that the valve at the points was operated by 
water in which salt or spirit was mixed, but in 1892 the 

present " Electro-Pneumatic " system was introduced, in 
which the points arc moved by high-pressure compressed air 
with valves controlled by electricity. 

As long ago as 1887 the Chemin de fer du Nord were 
operating their points by electric motors controlled by the 
ordinary Saxby and Farmer lever. In 1888 an installation 
of electrically operated switches on the Ramsey-Weir method 
was in use on the Cincinnati, Hamilton and Dayton RR.; in 
1893 the first Taylor plant was laid down at East Norwood 
on the Baltimore and Ohio RR., and in 1894 the Siemens- 
Halske " All-Electric " system was in use at Prerau on the 
Kaiser Ferdinands-Nordbalm, in Austria. 

The first low-pressure compressed plant laid down was 
designed by Mr. J. W. Thomas, now president and general 
manager of the Nashville, Chattanooga and St. Louis RR., 
and was fixed at Nashville in 1897. 

These plants were the forerunners of the several systems 
which, together with the original or parent schemes, are 
described in the subsequent pages. But, as in Mechanical 
Railway Signalling, only existing systems will be referred to, 
abandoned methods being ignored. 

The various methods in use are the Westinghouse 
Electro-Pneumatic, the Low-Pressure Pneumatic (" all-air "), 
the Bianchi-Servettaz Hydraulic, the Crewe, the Siemens- 
Halske, the Taylor, the Union Switch and Signal Co. 's, the 
Siemens Bros', and the Bleynie-Ducousso ("all-electric") 
and the Sykes Electro-Mechanical. 

Each of these possess in a more or less degree the advan- 
tages, together with the disadvantages, that are to be found 
in power signal plants. The advantages are many and great, 
and far outweigh the drawbacks, yet the latter merit con- 

Disadvantages of Power Plants. 

Signal boxes have been dispensed with and one cabin 
provided instead of two, which has led to certain of the 
points being further from the one box. This may lead to 
their being out of a convenient range for shunting opera- 
tions, and it may be necessary then to send a man to com- 
municate from the scene of shunting to and from the signal- 
box. A case is in the Author's mind where a power plant 
was fixed to replace two mechanical boxes in a busy goods- 
yard, but when completed it was found that the extreme 
points were so far away from the signal-box that a bell code 
had to be set up and a man employed near the points to 



signal to and from the box what movements had to be made. 
Consequently, the economy effected by saving a signal-box 
\vas swallowed up by the men necessary for the extreme 

Another point to be remembered about the cost of power 
plants is the extra expense necessary for generating the 
power. Whatever system is employed, be it all-electric, 
electro-pneumatic, or pneumatic, a certain amount has to be 
debited for generating the power. Engines, engine-houses, 
batteries, battery rooms, generators, air compressors, etc.. 
etc., have to be provided and men employed to look after 
them. \Yhcre the power can be obtained from some existing 
source, or where the power can be utilised for some other 
purpose, then the cost is reduced. For instance, an all- 
electric plant can obtain its power from an electric lighting 
plant, providing, of course, that the plant is of sufficient 

Another important consideration in connection with 
power plants is the cost of their maintenance, and this is 
bound to be high for wages. The working parts have not to 
stand the same strain as in a mechanical frame, nor are 
there any rods, rollers, cranks, roller and crank frames, 
signal wires, wire pulleys, or wire pegs to maintain and 
renew, but the serious results that might arise from a failure 
of the apparatus renders it necessary for a skilled workman 
to be always near. In mechanical signalling a platelayer 
can generally put a breakage right, or at least sufficiently so 
to keep the signalling going until the chargeman can come to 
repair it properly. But in the case of a power plant only a 
skilled workman can remedy a defect, and often the greatest 
difficulty is to locate the defect. 

It is a sore temptation, in order to economise signal- 
boxes, to sacrifice the signalman's view of the line. The use 
of " Track-Circuits " may mitigate this evil to some extent, 
as any train, or part of a train, standing on a line equipped 
with the " Track-Circuit " will hold the signals for entering 
on to that line at danger. But this does not apply to those 
cases on British railways where passenger vehicles with un- 
bonded Mansell wheels especially if unbraked are left on 
the line, and certain goods stock with iron wheels (or cut 
down Mansell centres) will not shunt the relay. Consequently 
there is a possibility remote, may be, but worthy of thought 
where a vehicle could be on the line equipped with the 
' Track-Circuit " and yet the signals applicable to that line 
could be lowered. Further, a signalman ought to see the 
trains arrive and depart, so that he may be in a 
position to give thi necessary block-signals after his own 
personal observation. This defect could be met by the use 
of an illuminated diagram, but this would not tell the signal- 
man whether the tail-lamps were on the last vehicle. 

A good view of the line is essential, and it is easv to 
imagine how a good case for substituting two or more 
mechanically-operated signal boxes by fewer boxes operated 
by power might have to be rejected because as good a view 
of the line would not be obtained under the new conditions 
as under the old. 

From an operating point of view there is a source of 
danger in power-plants owing to the greater ease with which 

points and often locking-bars can be " thrown " when 
vehicles are on them. When mechanically operated the sig- 
nalman can readily tell by the drag on the lever that some- 
thing is on the points or bar, but there is no such intimation 
where power is employed. Power-worked bars have often 
been lifted and carried over when wheels have been on them. 
Out of the different systems it is difficult to select one that 
is not without its drawbacks. 

The " Electro-Pneumatic " relies on two forces air and 

The " All-Air " necessitates a network of pipes and is 
slower in its movements than electricity, but is cheaper. 

The " All-Electric " calls for a greater consumption of 
power and it is difficult to trace defects, whilst trouble is 
likely to arise from extraneous currents. It is, however, 
quick and certain in its work. 

The " Hydraulic " is cheap so far as the generation of 
power is concerned, because sufficient power can often be 
obtained by hand-pumping, but there is constant trouble to 
keep it from freezing in winter. The Bianchi-Servettaz sys- 
tem is in general use in Italy, Southern France, and Spain 
all countries where there is no hard frost yet the Paris, 
Lyons and Mediterranean R., which runs through the mild 
climate of the South of France, has only three plants of from 
14 to 22 levers. That company reported to the International 
Railway Congress (London) of 1897: "The cost of main- 
tenance is high. The apparatus is delicate and exceptional 
care is required to prevent the freezing of the liquid in the 
tubes." Since then an installation has been fixed at the Quai- 
D'Orsay station, Paris, where it was used as the small 
amount of room available precluded the use of channel sec- 
tion rodding or double wires. The space was so small that 
copper pipes 12 mm (15/32111.) diam. were used instead of 
iron pipes 16 mm (fin.) diam. It is, however, found neces- 
sary to add 40 per cent, of glycerine to the water in winter. 
This expensive liquid cannot be used in summer, as the heat 
decomposes the liquid and gas is evolved. There is, further, 
slower operation at greater distances, on account of the 
elasticity of the column of water, which nearly always in- 
cludes air bubbles. 

The " Electro-Manual " is a compromise, and like all 
compromises is an easy but not altogether satisfactory way 
out of a difficulty. The operation of signals by electricity 
is a simple matter, it is in the movement of points that the 
difficulties arise. Disposing of signals only takes a small 
amount of work off the signalman the heavier work of the 
shifting of the points remains, and whilst no signal wires 
exist point rodding remains for which valuable space is 

Advantages of Power PJants. 

Turning now to the other side of the question, let the 
advantages of power plants be considered. 

A very prominent advantage lies in the reduction of the 
size of the signal-boxes, which are becoming unwieldy in 
size and consequently take up a lot of room which often 
cannot conveniently be spared, and, as a matter of fact, the 
permanent way has frequently to be re-arranged in order 
to provide space for the signal-box. Mechanical locking 



frames have grown to a tremendous size, and no matter how 
close the levers arc placed to each other, and the sequence 
of the levers arranged so that those applicable to a certain 
operation are close together, considerable time is spent by 
the signalmen in going from one lever to the other. 

In a power plant the levers or their equivalents are placed 
at less than half the usual distance apart, and a considerable 
reduction is thereby made in the length of the frame and 
consequently in the size of the signal-box, which may also be 
located in a more inaccessible position than an ordinary 
mechanical signal-box may be, as there are neither signal- 
wires nor point-rods to be considered. 

The working of facing point-locks with the switches can 
be safely done by power, and the selection of signals, i.e., 
the actuation of one of a set of signals by one lever the 
position of the points " selecting " the signal that may be 
lowered can, with advantage, be adopted with power plants, 
so the length of the locking frame can be still further re- 

Bv avoiding the use of signal-wires and point-rodding, 
much economy is effected, as the cables or air pipes, or what- 
ever is used in power installation, can be led overhead or 
buried in any direction, whereas ordinary point-rodding 
and signal wires have to be taken in the most direct route. 
There is also no painting of point-rodding and no removing 
when the permanent way is rclaid or altered. 

The Board of Trade, where power installations are 
adopted, have relaxed their requirements as to the maximum 
distance for facing points, which may now be 300 yards dis- 
tant from signal-boxes, and so for passing places on single 
lines of railway, where the points arc worked by power, 
it may not now be necessary to provide more than one signal- 
box, as the extreme facing points may be 600 yards apart, 
and, therefore, 500 yards or so from the fouling point at 
one end of the yard to the fouling point at the other. This 
is an increase of about 200 yards over mechanically operated 
and sufficient to hold a train of 70 wagons. 

In the laying out of busy stations, yards and other new 
works, engineers will find it a great boon not to be tied 
down to get facing points- within 200 yards of a signal-box, 
and signal engineers will welcome the concession as one 
helpful to them in planning their work. 

An advantage of very great moment is the absence of 
rods and wires for men working on the line to trip over, 
especially in view of the legislative action taken with regard 
to the boxing-in of rods and wires. Nor do the points and 
signals require adjustment. The signals always come "off" 
to the proper angle, whatever be the state of the weather. 

Another advantage is the elimination of mechanical 
detection. At the present time in mechanical signalling 
the wire of a signal for passing through a set of points has 
to be threaded through the points so that if the switches 
are not properly " home," or if the rodding has failed and 
the points have not been altered, the signal cannot be low- 
ered. This means a complication of rods, slides, weights, 
and wires at each point, and this complication is magnified 
if a signal is for a route over more than one set of points. 
There then comes trouble in adjusting the signal wires; a 

trouble that is hard to explain, and of a magnitude that only 
those can appreciate who have had to deal with the vagaries 
of a signal wire detecting two or three sets of facing points 
on a spring day, that is hot when the sun is out, but drops 
below freezing point during the night. All this is avoided 
where power is employed, as the full stroke of the lever 
working the signal cannot be obtained unless the switches 
are lying right and properly "home." Another frequent 
cause of complicated adjustment of mechanically detected 
signals which is removed by the use of power-worked signals 
is where one set or more of switches are on one side of the 
signal-box and another set on the other side of the box. 
The position or number makes no difference in power-plants. 

The plunger-bolt of facing point locks is also detected. 
In the latest requirements of the Board of Trade as to new 
railways, etc., they agree to lacing points being 250 yards 
distance from a signal-box if the plunger be also detected. 

Then, as "Track-Circuits" generally form part of the 
complement of a power-plant, trailing points, i.e., the points 
in the main line of trailing junctions, siding connections, 
crossover-roads, and, what is of the greatest importance, 
the safety-points of sidings and goods lines, are detected so 
that they must be in position before the main line signal may 
be lowered. 

The same generating station that provides power for 
operating a signal power-plant will often furnish power for 
operating several sections of automatic signals, e.g., the 
six miles between Andover and Grateley actuated from the 
Grateley power house and the intermediate automatic signals 
between Woking and Basingstoke. Power operated signals 
on the latter section are semi-automatic in another sense to 
that generally conveyed by the words " semi-automatic," 
i.e., they are returned to the "on" position independently 
of the signalmen's action. They are also absolute automatic 
signals being lowered and raised automatically by passing 
trains at those times when the signal-boxes (power-worked) 
are closed. 

Question of Economy. 

It is now necessary to deal with two points which arc 
often quoted as advantages in favour of power-plants the 
saving of time in operation and the saving of men. 

The operation of power-plants imposes no physical exer- 
tion on the signalmen, but in the Author's opinion this aboli- 
tion of physical labour does not lead to a saving of time. 
In fact, where more than three or four levers have to be 
pulled for an operation, the work (except in the Siemens- 
Halske and Siemens Bros, system as fixed at Derby and 
Didcot) could be done more quickly by means of a mechanical 
frame. This slower operation is due to one of the advantages 
of a power-worked machine. It is the " Return-Indication " 
which causes the slower result. One of the merits of a power 
machine is that the full stroke of the lever cannot be com- 
pleted unless and until the movement intended, and which 
is started when the lever is partly over, is completed. This 
is controlled by a lock known as the "Return-Indication," 
which holds the lever when partly over, and when the power 
has done its work a return current of electricity, air, water, 



or other force employed, comes back to the locking frame 
and takes out the lock and allows the full stroke to be 
completed. The merit of this is that should the power not 
have done its work, the lever cannot be pulled completely 
over, and so not only does the signalman know that some- 
thing is wrong but the holding of the lever prevents sym- 
pathetic point or signal levers being pulled. This " Return- 
Indication " takes time it may be only a question of seconds 
but it necessitates the signalman waiting to complete the 
stroke before he can pass to the next lever, as the first has 
to be fully over before the second can come. In the case 
of the Low Pressure Pneumatic, and the Taylor " All- 
Electric " the stroke is completed automatically, but the 
signalman cannot move the subsequent lever until the full 
stroke is automatically completed. It must therefore be ad- 
mitted that where several levers have to be moved for one 
operation they can be whisked over quicker in a mechanical 
frame than they can be moved in a power-plant. 

It is mentioned above that this objection does not apply 
to the Siemens system. In this the control-fields hold the 
signal concerned until all the point levers have responded, 
but the latter can all be pulled over as rapidly as possible, and 
the signalman move quickly from one point lever to the 
next. But a feature that is prominent in other systems is 
omitted, viz., should anything fail in the other systems the 
" Return-Indication " of the affected lever would not come 
in, and the signalman would at once know which lever was 
wrong. In the Siemens system the failure would not be 
known until the signal failed to come " off " and this might 
be due to one of the points concerned being wrong, or the 
fault might lie in the signal itself. 

Another doubtful point about power-plants is as to 
whether the number of men employed can be reduced. It 
is claimed that the substitution of a mechanical plant in a 
signal-box by a power-plant will lead to a reduction in the 
number of men employed in that box. This is not always 
so, in fact rarely is it the case. Take for example some of 
the large termini. In the signal-boxes there the locking 
frames are divided and a chief signalman and two or perhaps 
three assistants are employed. These have each their allotted 
tasks, one man taking the incoming and a second the out- 
going signals for one side of the station and the other two 
men take the incoming and outgoing signals for the other 
side of the station. In addition to working the locking- 
frame they have the block instruments to attend to, work 
which is the same whatever system of signalling is employed ; 
as is also the number of train and shunting operations and, 
consequently, the movements of point and signal levers which, 
as just pointed out, take up more time in a power-plant than 
in a mechanically operated, owing to having to wait 
for the "Return-Indication." The exchange of signals on 
the block instruments is not reduced, and the natural division 
of the work into east and west or north and south and " up " 
and " down " continues. Consequently it is not found ex- 
pedient to reduce the number of men employed, although it is 
admitted that their labour is lessened. 

It is the traffic movements and not the amount of work 
that is the governing factor. 

Economy is effected when signal-boxes can be concen- 
trated owing to the increased distance for facing points 
allowed by the Board of Trade for power-worked points. A 
fine example of this is at Staines, L. and South Western 
R., where two power-operated boxes do the work of the 
former five mechanical boxes, and with increased facilities 
in addition. The Guide Bridge widening, Great Central R., 
Cromwell Curve and Mill Hill Park on the Met. District R. 
furnish other such examples, whilst the whole of the enlarged 
Central station, Glasgow, has been signalled from one power- 
worked box. There is quite a fruitful field for economy in 
this direction and in some cases, to continue the simile, it is 
"ripe unto the harvest." A case in point are the nine 
signal-boxes between Exchange Station, Liverpool, and the 
Kirkdale Tunnel on the Lancashire and Yorkshire R., which 
could be replaced by five boxes, for which the electric power 
" passes the door," as these are on, or quite near to, the 
electrified Liverpool-Southport line. 

The Waterloo terminus of the L. and South Western R. 
is being enlarged and it will, no doubt, then be signalled by 
power. Part of Euston terminus, L. and North Western R., 
is signalled by the " Crewe " system. The recent enlarge- 
ment of the Victoria terminus of the L. Brighton and South 
Coast R. was an excellent opportunity whereby one power- 
worked signal-box could have done the work for which three 
electro-mechanical ones, with a considerable amount of cross 
slotting of points and signals, have been erected. In this 
case reliance on " Track-Circuits " would have been neces- 
sary for lines out of sight, instead of the present arrange- 
ment of relying on electrical fouling bars. 

It is certain in the future that no large station will be 
signalled mechanically. Some stations have, of necessitv, 
been and others will have to be signalled by power. The 
Mansion House Station, Met. District R., could not 
have been efficiently signalled mechanically, whilst demands 
of space and the difficulties of obstructed view prohibited any 
idea of mechanically signalling the Grand Central Station of 
the New York Central and H.R. RR., and the Pennsylvania 
RR. Terminus, both in New York. In fact, one cannot asso- 
ciate mechanical signalling with work of such dimensions and 

' Track-Circuits " for ensuring that the lines are free 
and the fouling points unobstructed are an essential feature 
in all these cases, as the signalmen have practicallv no view 
of the roads they control. They are being used in America, 
particularly on the Pennsylvania RR., instead of locking bars; 
the " Track-Circuit " section, when unoccupied, energises a 
magnet that holds a lock out of the point or signal lever. As 
soon as a train enters the section the magnet is de-energised, 
so that the armature on the lock is released, and it falls into 
the lever. Point levers that have to be reversed to release run- 
ning signals are held in the " over " position, and those that 
have to be normal are locked in the normal position. To 
guard against the inconveniences that arise when the road 
requires to be legitimately changed a hand-screw release is 
provided that takes the lock out, but only with a deliberate 
slow movement. 

The Author, however, does not like the abandonment of 



the locking bar. He admits that it is possible for the bar to 
be moved over when a vehicle is on it, especially when 
attached to the wide-headed loolb. rails used in America, but 
on the other hand the circuit might fail or be improperly 
operated. As an alternative a " Track-Circuit " should be 
used as an additional safeguard. 

An admitted advantage of the use of a " Track-Circuit " 
is that any length of line can be protected by it, a boon now 
that rolling stock is constructed with a wheel-base longer 
than the length of standard locking-bars. It can also extend 
beyond the point of the switch, but these advantages should 
not justify the abandonment of the use of locking-bars. 

"Track-Circuits" may be used, as at Staines in 
mechanical as well as power plants for " holding the road " 
when signals are a long way back from the points and junc- 
tions. Such an arrangement should be used with care, as it 
may prevent a signalman from legitimately altering his 

Where power is available it is an easy matter to work 
level crossing gates, and where power passes a mechanically- 
operated box it is also easy to lay it on to the signals, which 
at once acquire all the benefits of power-worked signals, for 
all that is necessary then is for electrical contacts to be 
placed on the lever in the locking-frame, so that when the 
lever is pulled fully over the circuit to the signal is completed 
and the signal is lowered. Immediately the lever is released 
in order to be put back, the circuit is broken and the signal 
returns to the "on " position. This is what has been done 
throughout the whole of the Met. District R. at those 
boxes which have not been equipped with power, and it is 
a matter of surprise to the Author that on the Lancashire 
and Yorkshire R. advantage has not been taken of the supply 
of electrical power on the Southport line to operate some of 
their mechanical signals. 

Great satisfaction is derived from the operation by power 
of distant signals that are connected to mechanical signal- 
boxes. As long ago as 1875 Mr. Sykes so worked the distant 
signals through Penge Tunnel, and except these and those 
on the North Eastern R. between Alne and Thirsk, referred 
to in Chapter XIII., and the up distant signal at Ledbury, 
Great Western R., which is 1,600 yards from the box and 
between which and the signal-box intervenes Ledbury Tunnel, 
there are very few such signals in this country. 

Some of the advantages to be derived from this method of 
working signals are that they can be any distance from the 
signal-box and consequently of the greatest service where a 
tunnel or viaduct intervenes or where a driver would get a 
better view by carrying the signal further from the box ; they 
come "off" to the correct angle, and go fully to danger; 
there are no wires to adjust ; no heavy weights on the 
signal or the back-tail of the lever for the signalman to 
lift and which, by the way, often cause accidents to men. 
For slotted (controlled) signals the system is admirable. The 
Author knows from experience the troubles associated with 
complicated slotted signals, such as a lower distant arm con- 
trolled by an upper, which in turn is actuated from two 
places. By the use of a power-operated signal the working 
is simplicity itself, as the work of each man interested is 

accurately done and there is no question of tight or slack 
signal wires, and men disputing as to who is to blame for 
the signal not coming "off" properly. And the driver al- 
ways gets a distinct signal and is no longer in doubt. 

Soon, very few distant signals in America will be 
mechanically connected to signal-boxes. Conditions are ad- 
mittedly different there as the signals are either coupled by 
rod or by double (return) wires, and in winter great trouble 
is experienced from snow and ice, and by extreme variations 
in temperature all the year round. 

Power signalling provides a ready means whereby points 
and signals can be controlled from other boxes, and some 
power systems are so constructed that if the points are run 
through, neither they nor the mechanism shall be damaged. 
In Germany this is a condition, whether points are power 
or mechanically worked, that if they be run through they 
shall remain in the position they are forced to by the train. 

In America there are also peculiarities which, after being 
mentioned, need not be further referred to. In order to guard 
against a sin which seems peculiar to that country and which 
is generally known as " taking his signal away " approach 
locking has to be provided so that when a driver has passed 
the distant signal or come within a certain distance of the 
signal-box the signalman cannot alter the position of his 
points. Another feature is the necessity for providing signals 
for " backing up " the facing road. 

Of recent years a craze has arisen in France for "itinerary 
levers," whereby the whole of a movement can be set up 
by the operation of one lever. The idea seems at first sight 
to be delightful, but it is full of complications and a very 
costly fad, as in addition to points being worked when set- 
ting up any route and some points may be concerned in 
several routes they must of necessity be capable of being 
worked independently. Further, a minor alteration in a 
scheme will probably upset the whole situation, necessitat- 
ing the re-arrangement of the point and signal connections 
and the remodelling of the locking frame. 

As a summary of the number of power plants in use or on 
order may be of service the Author is, by the courtesy of 
the various firms interested, enabled to give the following 

Sieniens-Halske (Dec., if)o6). 

Control apparatus be- 
tween station master 

and signal-box ... 25 

Locking frames .... 156 

Levers ... ... ... 8,040 

Point levers 3,3^5 

Signal levers ... ... 1,029 

Route levers ... ... 1,397 for 2,448 routes 

Point motors ... ... 3,832 

Signal motors ... ... 1,205 f r 2 , 2 43 arms or discs 

Siemens Bros. 

Way and Works Sidings, Derby ... 48 levers 
Didcot North Junction ... ... 38 ,, 

Snow Hill North 224 ,, 

Snow Hill .South ... 80 



Low Pressure Pneumatic. 




Railway. LocHn 


Points. Signals. 

Name of Station. 


, x 


Great Britain 



| 3 


'- a 


Granary Junction ... 

Great Eastern ( 

4- 2" 



-, c c 

; o -Z * 



5 * 




Lanes, and Vorks . i 

4t 126 


Tyne Dock .. 

I !' 5 

^ j 
114 173 

Ground Frame 
Gratelev Andover . . . 

1 36 17 ... 


19 72 46 

6 8 ... 





'Newcastle ... 
22 'London 


North Eastern ... 2 

Met. District 13 
Baker St. and Waterloo .. ; 
G. N., Piccadilly, and B. ... 5 
Charing Cross, Euston and 
Hampstead 6 
Caledonian i 

99 '78 
186 391 
191 283 
'5 23 
18 37 

27 156 

79 , 258 

Salisbury East 
,, Ground Frame 
Basingstoke East 
West ... 









1 8 64 

'5 64 
4 8 
16 60 
'3 68 




30 2 
30 2 

3 4 
40 4 

,, Ground Frame 
Barton Mill ... 





I 4 
n 32 



14 4 


Est, France ... i 

10 15 

Barton Mill to Hook... 


28 *rvt t KI i P 





n 40 

2 3 

22 4 



Hook to Winchfield ... 

- 21 



> 2 14 I 

12 48 

34 28 4 


1- Prussian State 14 

383 202 

Newnham sidings 

I I 

4 8 

2 2 ... 

'VVanne "... 

Winchfield to Fleet 



12 Myslowitz 

Bavarian State i 

10 4 

Fleet to Farnborough. 

c 16 



5 32 

24 2 


Farnborough ... 
Sturt Lane 

""* '9 





9 40 
5 20 

2 3 22 
12 4 



East Indian 2 

too 133 

Sturt Lane to Pirbright 



4 'Cairo 

Egyptian State 3 

76 ; 74 

6 2 


3 '2 



Brookwood to Woking 
Staines East 

18 15 


5 40 

26 28 ... 


4 'Sydney 

Queensland i 
New South Wales 2 

16 22 

125 225 

15 10 


6 32 


'9 3 

Ground Frame 

i 22 13 
2 1 



4 40 
3 8 





New Zealand 2 

64 80 

ClaphamJunc.W. Main 

I 3 

4 '2 


5 5 




5ooo ;ooo 

,, West Windsor 

18 13 

43 28 ... 

5 36 
3 84 



22 12 
46 13 

Indicates a Passenger Station. 

West London June 
,, Ground Frame 

23 15 2 
34 2 

8 48 

.3 '2 






Crewe System "All Electric." 

Ashburvs East 

21 13 2 
24 2O 

4 40 
8 1 52 

30 24 4 
37 30 7 

Crewe North Junction 266 

,, ' West ... 


I 4 ... 

12 48 

28 21 .. 

,, South Junction ... ... ... 247 

Priory ... 
Fairfield Ground Frame 

28 21 
21 8 ... 
22 II I 

3 ^ 

15 64 

7 36 
10 44 

T. 8 






,, Middle Sorting Sidings ... ... 152 
,, North Sorting Sidings ... ... 95 
,, South Sorting Sidings ... ... 76 

Audenshaw June. 

^ 12 

6 6 24 

16 10 


,, Salop Goods Junction ... ... 57 

Stockport June. 
Stalybridge June. 

5 20 

ii - 7 36 
19 ... 9 48 

28 21 12 

31 28 II 

,, Gresty 
,, Station 

^ane ... 



Cabin A 


5 6 

6 ... 8 20 

10 ... 12 36 

10 10 4 
18 18 4 

,, Station 

Cabin B ... ... ... 26 

7rt *>C 

17 12 

Huston No. 3 

_ 2 

JJc wsnsip ... 
Hyde Junction 

T 22 

9 - 20 
'5 7 44 

28 2 5 "3 

Huston No. 4 




7 5 24 

M 7 


Camden No. i 


ElsL-car June. ... 
M"<,r Road 
Wath B 




2 7 64 
... 6 32 
i 9 48 




Camden No. 2 
Severus Junctio 




n, York 

Wath C 
Brigittenau, Austrian 
Ermont, No. 2, Nord, 

' 3 




10 48 

... 13 28 



Electro Mechanical, " Sykes." 
Mechanical. Electrical. Total. 
Station, St. Enochs ... 88 400 488 




3 44 

Clyde Junction .. 

30 70 i oo 

Ermont, No. I 


10 ... 6 32 

Gorbals ... 

27 41 68 

_ o 

Port Eglington .. 

33 45 78 


3 s - CO 
O t^- 


o Shields 

12 18 30 

if, "i f ^ 


00 . 



6 16 22 

Victoria North .. 

21 77 98 

Victoria South .. 

106 163 269 

Shunting Box 

II II 22 

Grosvenor Road 

6 24 30 

Battersea Pier .. 

II 29 40 

Battersea Park .. 

21 49 70 







H'csiinghousc System. 

Ix this system the points and signals arc operated by air 
compressed to a pressure of about 70 lt>s. per sq. in., the 
valves being opened and closed electrically. It is the most 
largely adopted of all power systems. 

One of the largest installations of it is that at the Boston 
Southern Terminal Station, and which was laid down by the 
Union Switch and Signal Co., who hold the Westinghouse 
signal patents for America. 

The Boston Southern Station is the joint property of the 
Boston and Albany RR. and the five companies now merged 
into the New York, New Haven, and Hartford RR. It 
covers 35 acres ; the covered portion of the station is 6o2ft. 

In November, 1904, a census was taken and it was found 
that during the 24 hours there were 1,524 movements of 
trains exclusive of shunting, 5,953 routes were called, there 
were 3,093 through movements of trains and 19,459 signal 
movements and 12,465 point movements, making a total of 

Eight main line movements may be made at one and the 
same time. Of the 148 main line signals, 128 are mounted 
on nine iron bridges spanning the lines they apply to. They 
have red lights for danger, yellow for caution (where "cau- 
tionary " signals are used), and green for safety. Below each 
of the 28 signals which govern the departure of trains from 
the station there is a lower distant arm, which, when " off," 
indicates that all advance signals ahead are off. There are 
also " cautionary " signals for inbound trains, which are 
controlled by a " Track-Circuit " extending up to the buffers 
at the end of the station. These drop with the home signals 
above them if the road ahead is clear, so that a driver gets 
warning whether he may expect a clear road or not. 

o o 






Fig. 385. Electro-Pneumatic Dwarf Semaphore. 

in. length, and 5/oft. wide. The length of road under the 

roof is 4 miles, and the total length in the station and yard 

is 15 miles. There are 37 double slips, 252 single points, 

283 crossings ; under the station roof there is accommodation 

for 282 cars (each 65ft. long), and 740 trains use the station VERTICAL SECTION 

daily. The actual schedule number of trains using the station 

was 775 per day, and the average number of daily train 

movements 2,500. 



I'ig. 386. Signal Cylinder and Valve. Electro-Pneumatic Semaphore. 




V : 



2 Air 

^17 x K>* Point Tic's .- , 

fy-"- ^g TLJ=J< 

. ^___^ i " -- ~ "' 

_<i >*^*-"- -3^'- ^r^tj^ii X ^jf- 'f- v^rrH" J 'ii' ~<^' ^U 
' ."7 " ." ** ' r "v.\ ; v * T' C ". - ' c ----> . ,--' : r^t . . - 
"v* '.1 \*:-:V - -,>-/-'- '-*.< ^. 

"-" -. -o o i . o' * > ","- .*" -* ". ' <^ j jtrmoredHoj&Jl\ 


In tower (the American equivalent for signal-box) No. I 
there arc 143 levers, of which 130 are in work, and con- 
nected thereto are 91 home signals, 36 cautionary signals, 21 
dwarf siding signals, 31 double slips, 31 crossings, 49 single 
switches, or a total of 148 signals, and the equivalent 

of 233 single switches. 

The advantages of a power plant over the ordinary 

mechanical manually worked apparatus were very 
manifest here. To work the number of signals and 
points as already given would require a frame of 360 levers at 
least, which at 5in. centres would require a box at least iGoft. 
long. The lead-out for the point-rodding and signal wires, if 
arranged in the ordinary way, would take up as much accom- 
modation as would hold 67 cars (American), and, it is stated, 
the cost of a mechanical plant would have been greater to 
instal and also greater to operate. 

The advantages of the electro-pneumatic that were chiefly 
instrumental in securing its adoption were : It required only 
about one-third the number of levers, and a box one-fourth 
the size required for the mechanical, and permitted it to be 
placed on ground not required for other purposes, and in a 
position giving the best view of the switches and signals, 
whereas the mechanical box would have had to be fixed 
on one side. 

The box is worked by a staff consisting of a directing 
dispatcher and his assistant, a telephone attendant, a tele- 
graph operator, and three levermen during two of the day 
shifts ; at night a dispatcher, one operator and two levermen 
are required. The maintenance force consists of a general 
repairman and his assistant during the day, and an assistant 
repairman and helper at night. 

l<~jir. 388. Sectional Diagram Electro-Pneumalic Switch Cylinder and 
Valve (cover of valve removed). 

The air compressors for the plant are two Ingersoll- 
Sergeant piston-inlet machines, each having a capacity of 
382 cubic ft. of free air per minute, at a maximum speed, of 

Fig. 391 



Signal Lever in Normal Position. 
Signals at "Danger." Lever free 
to "clear" either Signal. 

Signal Lever in Forward Position. 
One Signal at "Clear." Signal Lever 
locked from being put entirely "Normal." 

Signal Lever in Semi-Forward 
Position. Signals at "Danger." 
Lever free to be put "Normal." 



I2o revolutions. The steam end of each compressor has a 
cylinder I4in. diam. x i8in. controlled by a double slide 
valve with the Meyer adjustable cut off. Only one machine 
is run at a time, they being used on alternate days. 

Each compressor is provided with a receiving tank with a 
safety valve. From these tanks a 3in. air main extends 
to the limit of the interlocking, and 2in. branches lead off 
in convenient lines, with jin. pipes from them to each point 
and signal. At each switch and signal there is a small 
auxiliary reservoir for collecting such moisture and sediment 
as may be contained in the pipes and carried along with the 
air. This insures clear dry air for each operating mechanism. 
The valves of switch movements are connected with these 
reservoirs by means of armoured hose, so that no pipe joints 
may be subjected to strains resulting from lifting and slew- 
ing the road, or from vibrations due to trains passing over 
the points. 

The cylinders for operating the points and signals vary in 
size according to the work they have to do. The sizes are : 


Dwarf signals 

High signals 

Single Switch 

Double slip end, no cross- 

Double slip end, w ith cross- 


Content* of 




cylinder (air at 

in Iree air 

lii-. In- 

sq. ins. 

90 Ibs.), cu. ins. 

cu. ft. 

3 2 




3 4 











The motor for actuating signals is the same as that used 
for the Union Switch and Signal Co. 's automatic signal, 
fig- 259. 

Fig. 385 shows the electro-pneumatic dwarf signal, which 
differs from the semaphore signal motor, the principal altera- 
tion lying in the replacement of the balance lever and weight 
by a spiral spring, thus enabling the signal to be of compact 
and light construction, and all the parts to be enclosed 
secure from the weather. 'I he cylinders are movable, and 
the pistons stationary, and are connected direct to the signal. 
It should be borne in mind that in this system there is not 
the objection to springs for putting signals to danger that 
there is in mechanical installations, as in the YVestinghouse 
method should a signal fail to go to danger, the full return 
of the lever in the locking frame cannot be accomplished, 
and the signalman thereby knows that the signal has not 
gone to danger. It will be seen on reference to fig. 385 that 
carried on the signal cylinder is a brass plate which, when 
the signal is at danger, closes a circuit by resting against 
two contact springs attached to the base of the signal. \Yhen 
the circuit is complete, an electric lock in the interlocking 
frame is withdrawn, which allows the lever working the 
signal to be put fully "home." (This will be seen more 
clearlv when locking frame is dealt with.) 

!"'& 3">5- Switch Lever in "Normal" Position. Dotted position of 
parts show preliminary movement of Lever made in reversing the 
Switch (switch not having yet responded). 

Fig. 394. Switch Lever in "Semi-Reversed" Position. Switch having 
moved and become locked in reversed position in response, and Lever 
having been electrically released in consequence for final movement to 
extreme reversed position. 

Fig. j<)2. Working E.P. Signals by "Selection."' 


Fig. 395. Switch Lever in "Reverse" Position. Indication latch 
released to again engage Lever when put "Normal" and again partially 

The semaphore signals are attached to signal bridges, 
and are constructed as seen in fig. 386. These are not pro- 
vided with springs, but the spectacles and back lights are 
made unusually heavy, and the signal rods of ijin. solid 
round iron, and so the return of the signal to danger is 
sufficiently provided for. 

The working of all these signals will be understood from 
the details given in figs. 385-6, coupled with fig. 259. 

P O W ER RAILW A 1' S / G N A L L I N C, . 

The connections for moving the switches are shown in 
fig- 3 8 7- Whilst the principle is the same as for the signal, 
the means for setting the mechanism in motion are different, 
and power is required for moving the piston both forward 
and backward, as points cannot be reversed by their own 
weight in the same way as a signal goes to danger. Points 
have also to be moved over in much less time than that in 
which a signal is pulled off, and the operation requires also 
much more work to be clone. For instance, with facing 
points the movements to be made are, first to withdraw the 
facing-point-lock, then move over the points, and lastly to 
bolt them again. 

Unlike general Hritish practice, all the mechanism in the 
electro-pneumatic system is placed outside the "four-foot," 
and is carried on timbers secured to the sleepers. The two 
switches are connected to each other by the rods a 1 a", the 
rod for operating them, a 3 , being coupled to the further 
switch, whilst the lock-rod a 4 is coupled to both switches. 
The switch rod a 3 is connected to the bell-crank b b l b~. This 
is not connected to the operating rod c, but works between 
the upper and lower bars c 1 c" which form the operating rod. 
The lock-rod a 4 has two slots cut in it, one, d l , on the lower 
side fitting a lug, d~, on the operating rod c and holding the 
points when they lie in the position shown in fig. 387. The 
other slot, tP, in the lock-rod a 4 engages the lug d 4 when 
the points are over. 

Air being admitted to the switch or operating cylinder, 
the operating rod c moves from right to left, and the lug d? 
is at once withdrawn from the slot d l and the lock-rod and 
consequently the points is freed. This amount of travel of 
the operating rod c brings the roller c into contact with the 
arm b 1 of the bell-crank which it pushes out of its path, and, 
by so doing, draws over the rod a 3 , and consequently the 
switches and the lock-rod a 4 . This movement also brings the 
other arm b- of the bell-crank behind the roller c and across 
the path of its return movement. The contact between the 
roller e and the arms b 1 and b 2 of the bell-crank lever b would 
have a tendency to bend the operating rod c out sideways, 
but this is prevented by the fixed roller i. 

The roller e passes free of the bell-crank arm b 1 and the 
movement of the latter therefore ceases, but lug cJ 4 comes 
up to the lock-rod 4 , and if the switches be properly home, 
the slot d 3 in u 4 will be in a position to receive the lug d 4 , and 
the movement of the operating rod will be completed. 

The operating rod c is continued to a crank (not shown in 
the illustration), the other end of which is coupled up to the 
detector bar / (" locking-bar "), which is of sufficient length 
to prevent it rising between the wheels of the longest vehicle 
in use, and is attached to the points, so that should any 
vehicle be standing on, or approaching, the switches, some 
of the wheels will be on the bar, which cannot therefore be 
moved. As this bar will start to rise directly the operating 
rod c is moved, it follows that should any wheel be on the 
bar the movement of the rod c is stopped, and consequently 
the points cannot be moved and a derailment is avoided. 

This arrangement can be extended so as to guarantee 
that trains standing on conflicting lines arc clear by fixing a 

bar at the fouling point and working it with the signal 
governing an operation passing over the fouling point. 

The points being properly over, it is necessary to indicate 
the fact to the signalman so that the signals for going 
through the junction may be lowered. In mechanical sig- 
nalling this is done by means of detectors which (to explain 
briefly) are slides attached to the switches, through which 
the signal wires pass, and are so constructed that if the 
switches be not properly home the wire for working the signal 
cannot be pulled, and consequently the signal cannot be 

In the electro-pneumatic system this is accomplished 
electrically. The lever or handle in the locking frame, after 
being moved sufficienth far to operate the points, cannot 
be moved completely over in consequence of a stop, which is 
withdrawn electrically when the switches have gone properly 
over, when the movement of the lever can be completed, and 

Fig. 396. Switch Lever and Connections adapted to operate a Cross- 
over by means of Electro-Pneumatic Movements. 

consequently the signals for going through the junction can 
be lowered. 

If for some reason the points have hung up and are not 
properly " home,'' or the lugs in the operating-bar have not 
entered the slots in the lock-rod, then the lever in the locking 
frame cannot be completely pulled over, and consequently 
the interlocking prevents the signals being released. 

The necessary indication is attained as follows : On the 
top of the upper bar of the operating rod c there is an arma- 
ture g, and when the rod c is properly over, it makes a con- 
tact between the spring switches g- g 3 in the indication box 
h when the operating rod is forward, and completes the cir- 
cuit, which withdraws the stop on the lever in the locking 
frame and permits the movement of the lever to be com- 
pleted and the interlocking for releasing the signals is thereby 

When the points arc reversed and the operating rod fully 



back, contact is made by the spring 
o- 4 g 5 , which allows the lever to be 
put fully back. 

The switch cylinder is illustrated in de- 
tail by fig. 388. It contains a piston 
a connected by the piston rod b to the 
operating rod c in fig. 387. Air from 
the main enters the chamber c 1 (in 
which the air pressure is constant) 
containing a slide-valve J, which con- 
trols the communication from the air 
chamber c 1 to the cylinder by either 
the port e l , to force the piston to the 
right, or by the port e- to force the 
piston to the left. The slide valve 
a also gives communication between 
the cylinder and the exhaust / by 
means of the ports e l e". The slide 
valve is held in position by the lock 
pin o- 1 , so that the first movement is 
to withdraw this pin. 

On the signalman moving the 

in the locking frame the 

magnet M 1 is energised 

and attracts the armature 

in 1 , which opens the lock 

exhaust g-, and thus reduces 

the pressure on that side of 

the lock-piston g- 3 . The air 

pressure on the other side then 

overcomes the spring inside 

the lock-piston g^, and forces 

the lock-piston g* along the 

lock-cylinder c-, and takes 
with it the lock pin g l , so 
freeing the slide valve d. In 

the illustration the piston a is 
shown over to the right, and 

corresponds with the position 
illustrated in fig. 387, and, 
therefore, to move the points 
over, the piston has to be 
forced to the left. Assuming 
that the signalman has made 
the first portion of the move- 
ment of the lever, and the lock 
pin g l has been withdrawn 
from the slide valve d, and the 
latter is free, the next move- 
ment of the lever will cause 
the magnet M- to be ener- 
gised, and to attract an arma- 
ture which raises the pin valve 
li 1 from its seat, and the 
end of the armature stem 
shuts the entrance to chamber 
c 3 (and the exhaust .v 2 ), and 




UMU umu 

Fig- 39^' The Westinghouse 
Electro-Pneumatic Interlocking Machine. 

allows air to circulate from chamber c l by means of the pas- 
sage s 1 into the smaller cylinder f , forces the smaller piston 
k, the stem of which pushes the slide valve d over to the left, 
so that port e 2 is connected to the air chamber c l . Air then 
enters the switch cylinder and forces over the piston a to the 
left. The air on the other side of the piston a is forced out 
through the port e 1 , which has been put in connection with 
the exhaust /. 

At the same time that magnet M 2 was energised, the 
corresponding magnet M 3 was de-energised, so that the pin 
valve 7i 2 was forced back on to its seat by the spiral spring 
I and thus stopped passage of air from the main chamber 
c 1 into the smaller cylinder f by means of passage s 2 , and 

at the same time opened passage s- to the exhaust :v 3 . The 
slide-valve d has now been pushed over to its left position. 
So long as the magnet M 1 is energised the pin-valve g- will 
remain open, and owing to the exhaust A' 1 the pressure on 
the lower side of the lock-piston g 3 will be less than the pres- 
sure on the upper side of it, the spring will be overcome and 
the lock pin held out of the slide-valve. Hut upon a further 
movement of the lever in the locking-frame, the lock-magnet 
M 1 is de-energised and immediately the pin-valve g~ is re- 
turned to its seat by the spring above it. The compressed 
air then passes through the passage o (which is smaller than 
the passages o 1 o 2 ) in the lock-piston and equalises the pres- 
sure on both its sides, and the spring inside the lock-piston 
forces the lock-pin into the other seat in the slide-valve. 

It will now be interesting to give a description of the 
electro-pneumatic (American) locking frame, and to examine 
the details of the well-arranged mechanism whereby the sup- 
ply of compressed air that actuates the switches and signals 
is governed, and to explain how the means whereby this is 
done, are adapted to, and controlled in the same way as an 
ordinary mechanical interlocking frame. 

Fig. 389 shows a lever that works either of two conflict- 
ing signals, one the "up" by a "push" movement, and 
the " down " by a " pull " movement. The lever is nor- 
mally in the centre and on being pushed back in the quadrant 
the signal magnet F is energised, and a current is sent to 
the " up " signal, which admits air to the cylinder as already 
described, so that the signal arm is lowered as seen in 
diagram fig. 390. Immediately the signal arm leaves the 
danger position the lock circuit is broken, and the magnet 
holding the electric lock becomes de-energised, and causes 
the lock a to fall on to the tappet b attached to the lever. 



Sbfe:*- Allmonments over rlorrtavti Dtfrsron Main Tracks retrv're 
ffre cooperation gf ' tererm^rno 

No te - Signals marked x are suspended 

^Thfse art Imaginary Jtgnats tfthich must 
^ precede all signals governing movements 
Eastward from Torver "A " 


1 Lerer lor Detector Bars. 

11 Lecers for S Siritches, S Double Slip* iri'/i Moreable Froye, and S pair 
Moteable Frogs, and 1 Double Slip trith Rigid frog. 

5 Lecers for 2, 1 arm high Signals and li Uirarj Siynali. 
17 Working Leten. 

Spare Lereri. 
1~ Lecer Machine 

77 Lereri for 5J Single Stritchei. il, Double 
Slips irith Morfable Frogs, and i Single Slips 
with Jlureable Fry*. 

55 Lecers for il Arm and 21 Arm high Sig- 
nals, 7J Dirarf and 10 Suspended Signals and S 
Imaginary Signals. 
li: Wurting Lereri. 

te Spare Lerers l-l-S-i-5-6-S-10-l!-U-Sl-Sl-se- 

167 Lecer Machine. 

Fig. 399. Plan of Signalling at Long Island City. 

\Yhen the lever is put partly back to normal, as seen in dia- 
gram fig. 391, the signal magnet F is de-energised, and" the 
supply of air to the signal is closed and the exhaust opened, 
so that the arm goes to danger by its own weight (as already 

In the description of the details of the operating 
mechanism, reference was made to a stop in the locking- 
frame preventing the full stroke of the lever being completed 
until the points were properly " home." There is a similar 
contrivance in the signal arrangements to guard against a 
signal not going fully to danger. It is not provided for the 
signal in the opposite, or clear, position, as no harm can 
arise from a signal not coming " off " properly. 

It has been seen that when the signal comes " off " an 
electric lock falls on the tappet, and it will be observed that 
there is space enough between the lock and the stops on the 
tappet for the lever to be moved some way. This movement 
is enough to put the signal to danger, and when the arm 
reaches the horizontal the lock circuit is again closed, and 
the magnet of the electric lock energised, which raises the 
lock free of the stop on the tappet as seen in diagram fig. 
391. The lever can now be put to normal, and it will be 

seen how, should the signal have failed to go to danger, the 
signalman is advised of the failure, and his further move- 
ments are stopped. The signal not being at danger locks up 
the apparatus. 

The method of working the down signal is similar, except 
that the magnet B is the controlling force, and the electric 
lock falls on the other side of the stops, and prevents the 
lever being pushed back to normal. 

The electro-pneumatic system is well adapted for work- 
ing more than one signal by the same lever, by what is tech- 
nically known as "selection," and which is illustrated by 
fig. 392. There are two levers, Xo. i for the signals and 
Xo. 2 for the points. The signal lever is fixed in the mid- 
way position, and will work either of the two arms B B when 
in the back position, or either F F when in the front position, 
and the position of the points determines which of each two 
signals shall be lowered. It should be remembered that the 
illustrations now being described are obtained from American 
sources, so that the signal arms are shown on the opposite 
side of the posts to that adopted in this country, and as the 
trains in America run on the right-hand line, the points 
as shown in the diagram are set for the main line, and 



therefore when the point lever is in its normal position, the 
upper arm B would be lowered when the signal lever was 
pushed, or if the lever were pulled, then the top signal F 
would be lowered. When No. 2 lever is operated the points 
are reversed, and the electric selector is moved over to the 
opposite magnet and couples up the two other arms. 

Coming now to the movements of a point lever. In 
diagram fig. 393 the point lever is shown in its normal 
position and the first movement causes switch valve magnet 
N to be de-energised, and the supply of air shut off from 
that side of the cylinder. The magnet L is energised, and 

over and bolt lock R inserted in the lock rod, and the indica- 
tor P having gone from N N to R R, a current is sent to 
the indication magnet R, which causes the lock to be raised 
.free of the tappet, and the lever can consequently be pulled 
fully over. In making this last movement a lug on the 
lower side of the tappet moves the indication shifter from 
R 1 to N 1 , which causes the indication magnet R to be ener- 
gised, and the lock to fall again into the position shown 
in diagram fig. 395. 

Diagram fig. 396 illustrates the arrangements necessary 
where two sets of points (as in a crossover road) are worked 

Upper army on all signals gorern tn right, and loicer arms govern to Hie left 

Lower arms on Bridges Xos. A and I, are Auto caution signal* indicating 
whether tin' train slied tracks are occupied onlii after the unner arms are 

Distant signals on Bridges Xos. 9, 10, 11, indicate when cleared that 
complete route has been lined to train shed. 
Brfd g 'e"N S " Br ' dye Xo ' 8 indieate or repeat the position of signals on 

Dwarf signals on Bridges Xos. 1 and 2 gorern in-lionnd xwitrhim, 
movements, and are to be: observed by trainmen from the north or rear side 

Dwarf signals on Bridge " A " over Tracks Xos. 1 and 2 are for out- 
bound switching movements and gorern to Kxpress Yard only 

Dwarf signals on Bridge No. 7 over Tracks Xus. .*/,. 3f, and 1,1 are fur 
out-bound switching movements and nnrern to Track Xo 3,1 onlu 

L A. 


Du-arf signal IftR is to be controlled by hand twitch " D" and will 
*l f Z "w, V'Of'ed Volition when witch "D" is lined for express yard 
lead. When switch "D' is reversed dwarf signal 16R will stand in the 
stop position and will be controlled bu the Tower only 
East bound signals on Bridge Xo. HI are to be norma, 

Top blatl 
of facing point s 

Machine lock 
over Track N 
and opposing d 
be set at prone 

Track seetii 
" B " and Xo 

20 Lcrcrs fitr /* s,ngle Switches, 3 Doubb Slip 
2e 56 Signals 
44 Working Levers 

!3_Spare spaces J. 2. ff. 49. SI. S3. f7. S8. S3. 
59 ierer Machine 

Worvabfc Frogs & 2. May cable Frogs 

9 Lerers for 10 Switches, 2 Double Slips n 
."/_ n 18 Signals 

f6 Working Lexers 
_7Spart: <Spacs '.2.3,4-. 20,2223. 
23 Lever Machine 

h noreabttFrogs&IPRMor. Progs 

the lock-pin (g in fig. 388) is withdrawn. The magnet R 
being energised, air is admitted to the switch cylinder, and 
.the operating rod is forced over. The lever is now in the 
position indicated by dotted lines in diagram fig. 393, and 
the stop on the tappet coupled to the lever has come in con- 
tact with the electric lock, so that the lever cannot be pulled 
completely over. This is the stop to which reference was 
made when describing fig. 388. But the points being over 
and bolted as already described, the magnet R is energised, 
and its armature, which forms the stop, is lifted clear of the 
tappet, as shown in diagram fig. 394, in which it is seen 
that the bolt-lock N has been withdrawn, the switches moved 

Fig. 400. Diagram of the Lines and Signalling at St. Louis. 

by one lever. There are two sets of motors, indicators, etc., 
and both must have done their work before the " Return- 
Indication " can be given, and thereby allow the signalman 
to complete his movements. 

The electro-pneumatic (American pattern) interlocking 
machine is illustrated by figs. 397-8. The ordinary locking 
frame levers are replaced by handles which are not pulled 
over as in a locking frame, but are turned to left or right. 
The signal levers which work " selectors " are in a midway 
position, and turn to the left to operate one lot of signals, 
and to the right to work another set. 

One of the details sometimes supplied with the electro- 



for To~er Vf J. 

Lo&s fa &S**gkS*ittte3. 8 Dovblt Slips nttif*v**at>tf Preys 
21 Lc r*rj I 

12. Spare Spaces 1,2.3, 

Dtcarf signal 16L is to &< eontnllfil by hand sicitch " B," and trill stand in tie 
proceed petition vlien tviteh "B" ,* lined for Tract Xn. 5S. When ttriteh " B" 
is recerted dirarf signal i6L irill stand in the stop poiition and if ill be controlled 
by the Tower only. 

t'a.-t bound Signals on Bridge Xo. 10 are to be normally clear semiautomatic 
and controlled from Totcers 1 and 3. 

TO/I blades on Bridge " D " Tract* Xos. 51, 53. 55, 56 and 57 are controlled 
from Totrer Xo. S and gorern their respeetire tracts to Bridge Xo. 10; they must 
loct the semi-automatic signal* on Bridge Xo. Id in the ttop position. 

Lover blades on Bridge " D " Tract* Xof. 51. 53. 55, 56 and 57 are automatic 
caution signals controlled by tract circuits, and gorern their respectice tracts to 
Bridge Xo. 10. 

Top blade on Bridge " D " Tract Xo. 53 gorerns to Tract Xo. 58. The lover 
blade gorerns to Tract Xo. 57, and must loct the opposing signal on Bridge Xo 
Id in the stop position. 

Top blades on Bridge Xo. 11 yorern to Tracts Xos. 51, 5i, 5Z, $6 and 57. Louer 
blades on Bridge Xo. 11 govern to Tract Xo. 58. 
'' " 

aes on rge o. 1 govern to rac o. 58. 

Top blades on Bridge '' C " Tracts Xos. 5i and 5i gorrrn to outbound Mer- 
chant* Mill, lie Made* i/,irern tn ont-bmnid Kail*. I.oirer blnile* goreni to all other 
routft. TO/I blade on Bridge " C " Tract Xo. 5S gorerns to Tract Xo. 5i Lover 
blade gorerns to all other routn. 

Machine locting to be to arranged that all signals gorerning movements from 
Tract X c/.M to Tract Xo. 5S in either direction can lie *et at procetd nt the tinne time. 


W fcM_-^ _ TO ._ ^^ 

r y S*?rS~M*sJ*ttMleSefS ^rnHanfttrfnys. 

_33_SparvSfx>cta -j *.'<.* 
~ -- - t **.**.< 

?is LKuMaAi,* 

Fig 400 (continued). 

pneumatic plant is a track model, whereupon the lines and 
signals are shown with movable pieces for each signal and 
points. These pieces are electrically connected to the return 
indication, and move with them, thereby showing the signal- 
man how the road lies and the state of affairs at the moment. 
As the drawing is fully described, probably no further 
explanation is necessary. 

'''> 399 ' s a diagram of the signalling from two signal- 
boxes at Long Island City. Signal-box A contains i-;^ work- 
ing and 35 spare levers, and B contains 17 working with no 
spare levers. In A 55 levers actuate 93 signals. The station 
is on the lower right-hand side. There are numerous com- 
plications. An exceedingly heavy residential traffic is dealt 
with all the year round and a big holiday and race traffic also. 

The St. Louis installation is, at present, the largest in 
America; fig. 400 is a diagram of the lines and signals. 

In the train-shed, or station, there arc 32 lines divided into 
two sections of 16 lines each. Immediately outside the station 
the lines divide east and west, and as access can be obtained 
to and from each of the 32 lines in both directions the lay-out 
and signalling are necessarily very complicated. 

But yet there are only three signal-boxes. Of these the 

most important is Xo. i, which controls the entrance to the 
station and all the lines from signal bridge Xo. 10 to signal 
bridge Xo. 11. It contains 215 levers, of which 33 are spare. 
There are several cases where more than one signal is 
actuated by the movement of one lever to left or right in 
fact, 79 levers work 194 signals. 

In Tower Xo. 2 there are 2 machines, one with 59 levers 
(15 of which are spare) and one with 23 levers (7 spare). In 
Xo. 3 Tower there are 35 working levers and 12 spare. 

There are 2 1 bridges of signals and the four-sided bridge, 
Xo. 5, at the central level crossing. The lower distant signals 
on bridges 3, 4 fall with the upper arms when there is a clear 
line into the station. They are automatic signals controlled 
by " Track-Circuits " in the station. The distant arms on 
bridges 9, 10, 1 1 indicate that these signals and the inter- 
vening stop arms are " off." 

It will be noted that at junctions there are only two signals 
an upper arm to the right and a lower arm to the left. Xo 
attempt is made to give a signal for each separate line into 
which a train may run. For instance, approaching the station 
from east to west there are three roads from each direction 
into lines 1-16 on the east side and lines 17-32 on the west 
side, and the signals for these lines are on bridges Xos. 3 and 




62 Levers for 18 Switches and SI D. Slips with M, Frog.*. 
,, ,, 65 Signals. 

Working Levers. 
17 Spare Lever* l-^-6-S-10-n-SO-S8-/,0-/^-U-70-7i-10S-llS-l 16 
119 Lever Machine. 
Levers 70-7^-108-111-110 used for locking between towers. 


62 Levers for K Switches, 39 D. slips with M. Frogs. 

55 8$ Signals. 

117 Working Levers. 

U_Spare Levers 27-29-36-5S-80-82-8G-100-10f,-106-110-lZ9-lSO-lSl. 

l$l Lever Machine. 

Levers S6-SO-8t-86-lOO-10.',-106-110 used for locking between towers 


11 Levers for 10 Switches and S D. slips with M.P. 

6 Levers for li Signals. 

17 Working Levers. 

JJSpare Levers 1-2-4-18-20-22. 

23 Lever Machine 

Levers 1,-18-HO used for locking between towers. 



8 Levers for 9 Switches and 1 S. Slip with M. Frogs 

1 D. Slip with M. F 
J_ 9 Signals, 
li Working Levers. 
Jl_Spare Levers 2-U. 
13" Lever frame. 

Lever 2 used for locking between towers. 

Fig. 401. Signalling at Union Station, Pittsburgh ; Pennsylvania RR. 



4. There are six dolls on each bridge, so that there is a doll 
for each line, ana the signals thereon lead to any one of the 
different roads to which access can be obtained. The arms 
suspended from bridges Nos. i and 2, and worked by 

lay-out and provide a continuous rail when crossed from left 
or right. 

Taking the signals for No. i line, and remembering that 
in America the trains run by the right and that the signals 

Fig. 401 (continued). Union Station, Piltsburg, Pennsylvania R.R. 

levers 78 L, 82 L, 86 L, 122 L, 126 L, and 130 L (L signifies 
that the lever is turned to the left), correspond to our " inner- 
homes," and not only "hold the road" but as there are 
signals for opposing movements on the same bridge they 
make a dividing line. 

This reduction in the number of signals adds necessarily 
to the complication of the interlocking, and it has been found 
that the 79 signal levers in signal-box " Tower " No. i are 
capable of providing for 1,827 different movements. For 
lever 190 L there are 14*1 possible movements. 

Fig. 401 is a diagram of the signalling at the Union Sta- 
tion, Pittsburgh Pennsylvania RR. The two larger towers 
contain respectively 131 and 119 levers. There are 14 lines 
in the station or " train-shed." 

Fig. 402 is diagram of the signalling arrangements at 
Thompson, Pa., Pennsylvania RR., and is given to illustrate 
how American RR. companies utilise their traffic facilities. 

It will be seen that there are five tracks, No. i (the lowest 
line on the diagram) is for southbound passenger, No. 2 is 
northbound passenger, No. 3 is southbound freight, No. 4 is 
northbound freight, whilst No. 5 is for freight also^and is 
used in both directions. Crossover junctions are liberally 
provided, and the "scissors" crossings are fitted with movable 
point frogs which have two advantages they shorten the 

point to the right, the distant signal is seen on the extreme 
left, under the automatic stop signal for the section in the 
rear. The outer-home signals for No. i line are operated 
by lever No. 12 R. The top arm a is for continuing on the 
straight and is taken up by inner-home 20 R. The second 
arm b, also worked by lever 12 R, leads through crossing 17, 
17 on to the facing line and along that line to the three arms 
worked by lever 18 R on the inner-home signals. Here the 
train may be turned back to its original line through cross- 
over 21, 21, or on to No. 3 line through crossover 25, 25, or 
to its original line through crossover 31, 31, or it may con- 
tinue on the facing line under special regulations or even go 
in a facing direction, also under special regulations, on No. 4 
line, gaining access thereto by crossing 25, 25 and 29, 29. 
The lowest arm c on the outer-home signal is a miniature 
arm indicating an irregular movement. It is for gaining ac- 
cess to the facing line Xo. 2 by the crossover 9, 9, or to the 
southbound freight line No. 3 by crossings 9, 9 and 13, 13. 
The inner-home signal Xo. 20 R for Xo. i line needs no 
explanation, and the next signals are those on the right on 
the outside, immediately past the points 31. These are auto- 
matic signals and stand normally " off." 

There are southbound signals for No. 2 line, which is for 
northbound passenger trains. The first is the miniature arm 




Fig. 402. Sigralling at Thompson, Pa. ; Pennsylvania RR. 

worked by lever 10 R. This is for setting back in any direc- 
tion. The next are the three arms worked by lever 18 R. 
The top a is for going from the northbound passenger No. 2 
to No. i line through crossing 21, 21. The middle arm is for 
passing on to No. 3 line through crossing 25, 25, whilst the 
bottom miniature arm is for proceeding in any direction 
except those just named. 

For No. 3 line there is a distant signal, then two outer- 
home signals worked by lever 8 R. The upper arm a is for 
continuing on the straight and is taken up by inner-home 
16 R. The lower arm b is for passing through crossing 11, 
n, and is then taken up by one of the three arms worked by 
lever 18 R applicable to No. 2 line already noticed. The upper 
arm a of signal 16 R is for continuing on the straight, but 
the lower miniature arm b is for proceeding in any other 
possible direction. 

There are corresponding signals in the opposite directions. 
It may be observed that the conflicting signals on the bridge 
carrying the inner homes are worked by the same lever, but 
with a motion to the left instead of to the right. This is a 
great saving in levers and reduces the interlocking. 

All the running lines are protected by " Track-Circuits," 
which play an important part in the movements on the facing 
line. The running signals are also semi-automatic, i.e., they 
cannot be lowered unless the track be clear, and they are 
thrown to danger as trains pass them. No facing point locks 
arc shown on the diagram, but they are provided in all cases, 
and are actuated by the same levers that operate the points. 

Fig. 4020 is a view of the tower at Thompson, Pa., and 
in the front are the inner-home signals for both northbound 
and southbound lines. The signals on the first bridge with 
the arms pointing to the right arc for northbound trains. 
Contrary to British practice the arms for the same line, even 
if for opposing movements, are not on the same post. This 
is because cast iron posts are used. Judging by the photo- 
graph it would seem as though the arms " clashed," but this 
is not actually so. (See p. 77.) 

The signals on the bridge in the distance are the north- 
bound advance signals (in the " off " position) and the south- 
bound outer home signals. 

The first installation in Great Britain was at Bishopsgate 
on the Great Eastern R. The locking frame contains i i 

Fig. 403. 
Signalling, Granary June., Bishopsgate, Great 

Eastern Railway. 


47 Lt*<er Framt, 

for 21 SiqnoU & 4 Slott . 

. 43 Point*. IF PL* 3 Bar*. 




Fig. 404. Plan of Signalling at Bolton, Lanes, and Yorks R, 


29. - 45. POINTS, 15. FR LOCKS f BARS, 



levers for working 21 signals and 4 slots, 26 levers for 43 
pairs of points, and 3 bars and 10 spare levers. The frame 
of 47 levers occupies a length of loft. Sins., and it is of 
similar construction to those used in America. Subsequent 
frames fixed in Great Britain are made to a type more in 
accordance with British ideas. 

A diagram of the lines at Bishopsgate is given in fig. 403. 

The next installation was at Bolton, of which fig. 404 
is a diagram. 

The signal-box (fig. 405), which measures 28ft. x i2ft., 
replaced two mechanically-worked boxes, as shown on fig. 
404. It contains a locking frame (fig. 406), having 83 levers, 
viz. : 44 signal levers, working 101 signals and 22 slots; 30 
point levers, working 43 points, 12 facing-point locks and 
bars, and 7 fouling bars ; and 9 spare levers. One of the 
features of power signalling, viz., the compactness of the 
locking frame, is well illustrated in this instance. The levers 
are spaced at 2|ins. centre to centre, as against sJins. in a 
mechanical frame ; so that the electro-pneumatic frame at 
Bolton is only i8ft. long by 3ft. 6ins. wide. Two mechanical 
frames having a total length of nearly 56ft. would have been 
necessary to accomplish the same work under ordinary con- 

Fig. 407 is a view of a locking-frame (not that at Bolton) 
with the case removed. 

Fig. 405 is an exterior view of the signal-box, from which 
it will be seen that the entrance is through the ground-floor. 
It is one of the advantages of electro-pneumatic power sig- 
nalling that the ground floor is left quite clear, as is shown 
by the view of the interior of it, fig. 408, after providing 
room for the staircase, battery-cupboards, lavatory, lockers, 
and w.c. , there is still ample clear space. 

The compressed air, at a working pressure of from 55 
to 65 Ibs. per sq. inch, for operating the signal and point 
motors is obtained from two large compressors, fig. 409, 
made by Walker Bros., Wigan, and which are located in 
the railway company's electric light station not far away. 

These compressors have steam cylinders 7ins. diam. by 
iSins. stroke, and air cylinders of Sins. diam. by iSins. 
stroke, and are capable of compressing 70 cubic ft. of free air 
per minute. They were purposely made much larger than is 

Fig. 405. Bolton West Box, L. & Y. R. 

necessary for the signalling installation, because a supply of 
compressed air was required for other purposes. They only 
work about 10 hours out of the 24. 

The compressed air is led through a surface condenser 
before it enters the main, so that the amount of water which 
enters the pipes is exceedingly small. The three or four 
reservoirs close to the power house are blown out every 
morning and the others every week. 



Fig. 407. Westinghouse Electro-Pneumatic Locking Frame (with case removed). 

Fig. 406. E.P. Locking Frame in Signal Box at Bolton. 

Fig. 408. Ground Floor of E.P. Signal Box at Bolton. 


Kig. 409. Air Compressors for Bolton Electro-Pneumatic Signalling. 

The electric power for energising the magnets which 
operate the valves of the pneumatic motors, and for indicat- 
ing the movements of the points and signals to the signal- 
man, is obtained from accumulators situated in the above- 
mentioned electric light station. A current of 3'6 amperes 
at 14 volts is sufficient to work the whole installation. 

The cables are laid in wooden trunking Riled with pitch, 
which has been used in preference to bitumen because it 
allows the cables to be got at more easily should alterations 
be necessary. The cables are accessible in six or eight places 
where they are run through test boxes, which are marked 
by a small plate showing the word " Test." 

The air pipe, 2ins. diam., is also run in wooden trunking, 
which, wherever possible, is placed about I2ins. above the 
rail level, so as to be easy of access. The pipe is so laid 
that in every case the air has two ways of getting to the 
branch pipes leading to the motors, and any of the motors 
can be cut out without affecting the rest of the installation. 
For this purpose special non-seizable cocks, having their 
coned plugs inverted, arc inserted in the branch pipes. 

The signal motors are 3ms. diam. x 4'ms. stroke, and 

are illustrated by figs. 410-11. In the diagram, fig. 410, a 
is the air cylinder containing the piston b, which is connected 
indirectly to the signal arm by the joint c. Compressed air 
from the main is admitted at d, and passes to the top of the 
piston through the valve /, which is opened or closed by the 
electro-magnet m, to the core of which the valve / is attached. 
The valve / has two seats (not rigidly connected), and 
when the lower seat is closed, as in fig. 410, gravity main- 
tains the signal-arm in its normal or " on " position, and 
the top of the piston is open through the upper valve seat 
and the exhaust passage e to the atmosphere. When the 
signalman manipulates a lever he sends a current through 
the electro-magnet and thereby closes the upper valve and 
opens the lower one, and thus admits compressed air above 
the piston and the signal-arm is pulled down or to the " off " 

Fig. 411. Electro-Pneumatic Signal Motor. 

Kig. 410. 

Kig 41 1. Electro-Pneumatic Signals, Bolton. 




Fig. 414. 

The "calling-on" arms and both the one and two-arm 
dwarf signals at Bolton are provided with special motors hav- 
ing cylinders 3ins. diam. x 2ins. stroke. 

Figs. 412-13 arc front views of signals, and fig. 
414 is a back view, from which it will be seen 
that the motors on the post are 
very inconspicuous. The one-arm 
dwarf signal is shown in fig. 415 
and the two-arm in fig. 416. The 
arms are locked, so that they can- 
not be pushed off by anyone pass- 
ing, and the lock is not released 
until the air is admitted to the 
motor. Another feature of them is 
that the ends of their arms are 
made of india-rubber, so that 
should a shunter run into them he 
does not hurt himself or damage 
the signal. 

All the signal lamps are lighted 
with 8 c.p. electric lamps. 

In the Bolton installation all the 
circuits of the signals governing 
facing points are led through elec- 
trical contacts at the points as 
well as through the electrical con- 
tacts on the point lever, thus mak- 
ing it impossible for the wrong 
signal to come "off" should the 
points have been tampered with 
since they were last moved. 

The circuits for operating the 
motors on the distant or controlled 
signals are in every case led 
through a contact breaker, fig. 
417, attached to the home or 
controlling arm, so that it is im- 
possible to get the controlled arm 

off unless the controlling arm be 
first "off." These contact breakers 
are seen at the back of the arms of 
the signals in fig. 414. 

Fig. 418 illustrates the ease 
with which electro-pneumatic sig- 
nals can control or be controlled 
by mechanical signals worked 
from an adjacent box. The cir- 
cuits are here given by which the 
inner distant and the outer distant- 
signals, No. 70, are worked from 
Bolton West Box, and are con- 
trolled by contact makers on the 
mechanical arms which are above 
them, and which are worked from 
Bolton down box. 

In fig. 419 the two inner distant 
signals No. 21, and outer distant 
No. 21, belonging to Bolton Up 
box, are worked by electro-pneu- 
matic motors by means of a con- 
tact maker on the mechanical lever 
No. 21 in Bolton Up box, which 
supplies the necessary electric cur- 
rent. This current passes through 
contact makers on the electro- 
pneumatic signals No. 15, pull A, 
or No. 15, pull B, so that one of 
these signals must be in the 
" clear " position before the cor- 
responding distant-signal can be 
lowered. In the same way the 
outer distant-signal, No. 21, is 
controlled through a contact 
maker on the electro-pneumatic 
signal No. n, worked from 

Fig. 415. Electro-Pneumatic Dwarf Signals, Bollon. Fig. 410. 


Fig. 417. Electro-Pneumatic Contact Breaker. 

Bolfon West Cabin 
Electro - Pneumatic 


air pressure from the small cylinder c 1 and opens it to the 
exhaust, and the latter closes the exhaust of the small 
cylinder c- and admits the air pressure to it. The slide valve 
S is then forced over and the port p l opened to the exhaust e 
and the port p- to the air pressure ; the main piston P is 
forced to the other end of the main cylinder C and the points 
are reversed. The slide valve S is provided with a spring 

As air can at all times enter cylinder C the piston P is 
kept continuously in the position to which it has been moved 
and the switches held " home." Should the points be " run 
through," the air is simply forced back into the main and 
then when the vehicle has cleared the points the switches 
are automatically restored without any damage being done 
to connections. 

Fig. 421 shows one of the facing points together with its 

1 indication 


Latch Contact 


The withdrawal of the plunger breaks 
trical contacts in the indication-box or det 

or r 


ii \ 


I L 

n t ' 


1 Worked from Mechanical Cabin 
1 M n Electro-pneumatic Cabin. 

U B> 1 


Bolton Uonn Cabin p)7D 
Mechanical (| j^ 

70:'nd>c&t'0n |([ 

K\ ' ' 

Fig. 418. Electro-Pneumatic Control 

Bolton \Vest. It will be noticed that the indication circuit 
is passed through all the signals numbered 15 and 21, so 
that they all must be at " danger " before the release locking 
is effected on No. 15 lever in the electro-pneumatic frame. 

The point motors are illustrated by figs. 420 to 423. The 
mechanism operating the switches is attached to the piston- 
rod at J. The piston P is shown in its normal position, 
and is being held by the compressed air entering the cylinder 
C through the port p l . The slide valve S is also being 
held clear of the entrance to p l by the air pressure in c l , 
the pin valve Y 1 being held open by the magnet M 1 . When 
the point lever is moved to reverse the points the electric 
current is cut off from the magnet M 1 and sent to M 2 . This 
moves the pin valves V 1 and Y 2 . The former cuts off the 

Doirn Passenger 
Loop Distant 

are unlocked and locked. It will be noticed (fig. 421) that 
the covering lids on the 4ft. wav are very neat and give 
complete protection against the mechanism being damaged 
by long coupling chains, and also against shunters and others 
being tripped up. The lids are hinged at the sides and meet 
in the middle so that they can be made strong without being 
too heavy to lift, and when open give full and ready access 
to the mechanism. This arrangement was designed by the 
Author when he was signal engineer on the L. and Y. R. 

Fig. 422 shows a set of trailing points. Both switches 
are detected by means of electrical contacts in the indication- 
box. The connections are shown in the illustrations. 

The new E.-P. signal-box was opened without a hitch on 


Worked from Eleclro-pneumatic Cabin 
,. u Mechanical 'Cabin by E.P.Mctor 

Fig. 419. 
Electro-Pneumatic Control. 

15 Push B 

is Pull B. 

IS Push 'A m j% i/5 Pull A. 

IS Push B 21 IS Pull B 

BoHon West-Cabin _ 

15 Pull B\ | 

15 Push B , | 

IS PIS/I A 21 15 Pull A 

Hjlja K>funH 

J Ul ,.^... J ... -,t. 

15 Push A 

// Indication . 

,, / 

Botton tip Cabin 


"Latch Contact 


-Latch Contact 


7 Threads per inch- 

Fig. 420. Diagram of E-P. Point Motor. 

September 27th, 1903, and one of the old cabins was imme- 
diately pulled down, and within 24 hours a line (see fig. 404) 
was laid over the site of it. 

Owing to the many alterations in the yard, and to the 
level of the lines being lowered 2ft. gins., the formation was 
very loose, and resulted in the frequent moving of the rails. 
This, of course, interfered with the circuits, and thereby the 

4200. Point Motor (British Practice), Westinghouse System. 

signalman was at once notified of any such movement a 
most valuable feature. 

All the electro-pneumatic apparatus was made by the 
Westinghouse Brake Co. at their London works, and was 
erected by the railway company's own staff. 

On the North Eastern R. the electro-pneumatic system 
was first installed at Paragon Station, Hull, where there are 
two signal-boxes, containing 153 and 179 levers respectively, 
and at Tyne Dock, where there are five signal-boxes contain- 
ing a total of 164 levers. 

One of the interesting features on this work is the opera- 
tion of four simultaneously acting gates by power. They are 
controlled in a similar manner to points, small levers of the 
usual dimensions being employed to open and close the valves 
of the motors for working the gates and gate-stops. The 
gate motors are larger than usual in order to provide power 
for moving the gates in windy weather, but by a special 

Fig. 421. Facing Points and E-P. Motor and Detector. 



Fig. 422. .Trailing Points \vilh Electro-Pneumatic Motor. 

valve, controlled by the gate lever, the movement can be 
graduated to suit weather conditions, etc., and the gates 
can be quickly stopped or reversed. 

The whole of the signalling on the Metropolitan District 
R. is on the electro-pneumatic principle. The automatic sig- 
nalling was fully described in Chapter XIV. 

Ihere are 13 electro-pneumatic signal-boxes on the Met. 
District R. Of these that at Earl's Court East, shown in 
fig. 424, will give an idea of the conveniences of power sig- 
nalling, The box is carried on girders over the lines, and 
the cables leading to the points and signals are laid down 
the side-walls. (See p. 238.) 

Some of the boxes are situate in most inconvenient posi- 
tions for the signalmen to see the trains and the points and 
signals they work. Such a case is that at the Mansion House 
stations very busy box. But the difficulties, and dangers, 
are overcome by the use of an illuminated diagram. One of 
these is illustrated by fig. 326. 

In fig. 423 is a photographic view of the connections to a 
pair of facing points, as connected up on the District R. 
They differ to those at Bolton (see fig. 421) in the way the 
plunger and locking bar are driven. Instead of the plunger 
and locking bar being worked by the same crank, and the 
bar becoming disconnected without notice, the bar is 
coupled to a rocking shaft driven direct by the 
point motor, and the plunger is attached to the bar. 
The indication switch (seen outside the positive power rail and 
in front of the point motor), is coupled to the plunger, so that 
the whole works in sequence, and should any part fail, the 
indication switch will not send the "return-indication." 

The wooden boxing over the facing point locks on the 
District R. are specially made so as to dovetail into and 
interlock each other. They cannot therefore come loose by 
accident and become a source of danger to men walking on 
the line. 

All signals on the District R., even when operated from 

Fig. 423. Electro-Pneumatic Facing Point Lay-out. District Railway 



signal-boxes, arc put to danger automatically by means of 
the " Track-Circuits." Those worked from signal-boxes are 
controlled by the track circuits ahead of them, and the lever 
is provided with a check-lock of the " return-indication " 
type whereby the signalman can put the lever sufficiently 
far back to put the signal to danger, but the full stroke 
cannot be made until the train has passed over the protected 

The largest plant in Great Britain, except Glasgow, is the 
2ii-lever pneumatic frame at No. 3 box, Newcastle-on-Tyne, 
and which was erected in connection with the construction 
of the King Edward's Bridge. It is one of five new boxes on 
the same system. 

Fig. 425 is the signal diagram of No. 3 box, and in it 
many interesting features may be noticed. 

King Edward's Bridge lies in the direction towards which 
the four lines point that arc on the upper side of the diagram. 
These four lines are all for passenger traffic. There are also 
four lines towards Carlisle, but the upper two are goods 
lines. At the entrance to the station there are two fine 
bridges of signals, and outside the junction from Carlisle 
and from the South there are other bridges of signals carry- 
ing the up and down inner home signals, and, further out, 
there arc bridges carrying the outer home signals and the 
starting signals for the opposite direction. 

A feature of interest is the extensive use of " calling-on " 
arms, which is the more interesting as they are operated 

Fig. 425. 

Plan of Lines and Signalling at King 

Edward VII. Bridge, Newcastle-on-Tyne, 

North Eastern Railway. 


S3 Levers Working Signals 
6 " " Points & facing Pomt Locki 

B " " Up Line Inoficators 

8 " " Down " " 



section. This " holds the road " and the non-return of the 
signal lever to normal holds conflicting levers. The signal- 
man* must also put the lever fully back before he can get the 
signal off again. Fouling points on sidings and on diverging 
and converging roads are protected by "Track-Circuits." 

At Earl's Court Station, when the L. and North Western 
engines that have brought their trains from Willesden are 
exchanged for electric loco-motors and vice versa on the 
return journey, " wrong-road " working has to be resorted 
to by means of special interlocking between the East and 
West boxes. 

At Putney Bridge Station, the up-distant signal is on the 
L. and South Western R., and is a lower arm on one of 
that company's signals. It is, therefore, a mechanically 
worked arm, and it is lowered by being connected to a snatch 

The signalling on the Met. District, the Baker Street 
and Waterloo, the Piccadilly and Brompton, and the Charing 
Cross, Huston and Hampstead Railways was carried out 
by the Underground Electric Railway Co., the material 
being supplied by the Westinghouse Brake Co. 

by the same lever as the upper arm. The first half of the 
stroke of the lever lowers the " calling-on " arm and the 
completion of the stroke lowers the upper arm too. By the 
use of these signals intimation is given to a driver as to 
whether the line be obstructed or clear. By coupling both 
arms to one lever a large number of levers is saved, and 
also a considerable quantity of interlocking. 

" Selection " i.e., the connection of two signals to one 
lever, only one signal being free and that the one for which 
the points are " set " has also been adopted, but not so 
freely as opportunity presented and for this the railway com- 
pany and contractors are to be commended. Selection is a 
source of considerable economy, but it is the better plan to 
provide a lever for each signal, as then the signalman cannot 
so readily lower an incorrect signal. 

What has struck the Author as the most remarkable 
feature in this installation is the coupling of certain signals 


Newcas tie-on- 1 "vne and King Edward VI I 
Bridge, North Eastern Railway. 

The whole of the lines are equipped with " Track- 
Circuits," which will control the signals, and an illuminated 
diagram has been fixed in the signal cabin, together with 
Train-describers, operated by levers in the locking frame. 

There are 211 levers in Xo. 3 box, and the interlocking 
is arranged under the floor. 

In the whole of the installation which excludes the No. 
i box there will be 401 working and 62 spare levers. The 
former will operate 352 signals, 108 trailing points, 67 facing 
points and locks, n facing point locks, 8 clearance bars, 9 
train describers and 16 indicators. 

The work was carried out by McKenzie and Holland, 
Ltd., for Mr. Chas. A. Harrison, chief engineer of the 
northern division of the North Eastern R. 

The installation at Glasgow Central is described in 
Appendix C., p. 336. 

Messrs. McKenzie and Holland have fixed three large in- 
stallations of electro-pneumatic signalling on the Egyptian 
State R. at Cairo. One contains 47 levers 17 for working 
39 signals, 25 for working 43 points, 30 facing point locks, 6 
fouling bars and 2 indicators and 5 spare. The second con- 
tains 35 levers 17 for working 25 signals, 15 for working 


and facing-point locks to two levers, either of which will 
work them. The object for this is that when a signal, or a 
facing-point lock, has to be worked for trains from either of 
two directions, considerable time is saved by reserving a 
lever in the order of levers taken by the signals from these 

As an example : The outer home signal for the up pas- 
senger line from Carlisle has to be preceded by signals Nos. 
115 and 116 or 117 and 118, when No. 109 points are " over " 
and by one of Nos. 197, 198, 199, 200 and 201, and one of 
Xos. 202, 203, 204 when Xo. 109 points are normal. There- 
fore to save the signalman labour the outer home can be 
operated by No. 119 when " led " by No. 116 or 118, or by 
No. 207 when " led " by Xo. 202, 203 or 204. Near by 
is an example of the same consideration as regards the 
working of facing point locks. The point levers to be 
pulled when a train is crossed from the up passenger line 
from Carlisle to the up goods line are Nos. 37, 39, and then 
No. 38 will bolt No. 39 facing points. When the road is 
normal the points arc bolted by Xo. 206, which is convenient 
to the signals for the up passenger lines, as has just been 

Fig. 425 (continued). 

26 points and 22 facing point locks and 3 spare. The third 
is a small frame of 7 levers for 10 signals and 4 levers for 7 
points and 6 facing point locks. 

Fig. 426 is a diagram of the lines, signals and points. 

The same firm have fixed a larger installation on the East 
Indian R. at Howrah. The Station box contains 67 levers, 
of which 24 work 64 signals and 15 slots, 31 levers work 60 
points and 52 facing point locks, 3 levers are for point con- 
trols and 9 are spare. The box working the junction with 
the Bengal Nagpur R. contains 29 levers for operating 55 
signals and 4 slots and 18 levers for 40 points, 32 facing 
point locks and 6 clearance bars. See fig. 427. 

Three installations, of 5 locking frames working 205 
points and 327 signals, have been or will be fixed by McKenzie 
and Holland, Ltd., in Australasia. 

The Brake Co. have fixed an electro-pneu- 
matic plant at St. Lazare Station (C. de f. de I'Est), in 
Paris, and six plants containing 14 locking frames on the 
Prussian State R. and one on the Bavarian State R. The 
frame at Cottbus on the Prussian State R. contains 90 levers. 



Fig. 426. 
Plan of the Signalling at Cairo. 

A Box 

17 Levers for J9 Siqnak 
ZS " " 43 Points 30 r.PLoclls.S noting Bars* 2 Indicofors 

S_Spere Spaces 
47 Lever Frame 

Siemens-Ilalske Electro-Pneumatic Gates. 
Where road level crossings are protected from the rail- 
way by barriers, as in America and on the Continent, it is 
a simple matter to operate them by power, as they only 
require raising and lowering and have not to be swung, as in 
Great Britain. Siemens and Halske have an arrangement, 
fig. 428,. whereby the barriers are moved by air or some fluid, 
and the valves controlled electrically from a signal-box or 

Lillooah Siding to Salt Golahs 
Lillouali Siding to Shunting Line 
Lillooah Siding to Down Goods Line 

- E.I.R, Down Main to Down Main 

-- E.I.R. Down Main to Shunting Line 

>9ff^ ~~~ -- Down Goods to Down Engine Line 

Down Goods to B.N.R, Down Main 

^^&-2, - D(llvn Goods to Shunting Line 

&~^ "~~-^ ~ - -_ Down Goods to Down Goods 

( ^ ^ "" E.I.R. Enjine Line to Down Engine Line 

"^^ "-^ . E.I.R. Engine Line to B.N.R. Down Main 

- E.I.R. Engine Line to Shunting Line 

-. E. I.H. Engine Line to Down Gooilb 

The supply of pressure fluid to the working cylinder i 
is controlled by two double valves 5, 6, actuated by electro- 
magnets, 3, 4. The barrier beam is actuated by the piston 
through the piston rod 7, connecting link 8 and crank 9, 
fixed to the barrier beam. Its centre of gravity is so 
arranged that in the half-closed position it is situated ver- 
tically above the journals of the beam. The barrier has, 
therefore, a tendency to move automatically into the one or 
other end position as soon as it is moved in the one direction 

Up Main to Up Main 
Up Engine Line to E.I.R. Up Goods 
Up Engine Line to B.N.R. Engine Line 
Up Engine Line to B.N.R. Engine ( ine 
Up Engino Line te B.N.R. Up Main 
B.N.R. Up Main to Up E.I.R. Goods 
B N.R. Up Miin to E.I.R. Engine Line 
B.N.R. Up Main to B N.R. Engine Line 
B.N.K. Up Main to B.-N.K. Up Main _ 
Shunting I ine to Lillooah Siding ~.^ 
shunting Line to E.I.R. Up Main ^ 
Shunting Line to E, I.'R. Up Goods _ ^ 
Shunting Line to E.l.K. Engine Line __ 

Shunting Line to B.N.R, Engine Line 

Shunting Line to B.N.R. Up Main 

E.I.H. UpGoods to Lillooah Siding 

E.I.R. Up Goods to Up E.I.R, Goods --- 
E.l.K. UpGoods to E.I.R. Engine 


E.I.R. Up Goods to B.N.B. Engine .,' 

E.I.R, Up Goods to B.N.R. Up Main ... 

evers for narking 50 Srpna/s 

IS tOPoini-s. 32 F 

Point Locks 6 Clearancf Bars 
Q7 Lever frame 

30 f. /. R. oomv GOODS 

fig. 427. Plan of the Signalling at Howrah, East Indian Railway. 



or the other out of the middle position. For giving warning 
of the intended closing of the barrier there is connected with 
the actuating gear an audible signalling apparatus which is 
operated by a descending weight. The audible signal con- 
sists of an escapement wheel u which is connected to a 

at every double oscillation of the lever. For securing the 
signal apparatus in the wound-up condition a lateral arm 15 
engages in a corresponding notch of the armature 16 of the 
electro-magnet 17 when this is not energised and thus pre- 
vents any motion of the lever and striker. Two cranks, 

Fig. 426 (continued). 

B Box 

/7 Levers for 25 Signals 
15 26 Points S 22 F P Loch 
32 Working Levers 
3 Spare Spaces 10. 26. Z7 
~37 Lever Frame 

weight 10 by means of a toothed rack and pinion gear 12, 
and of a pallet lever 14 carrying the striker 13 of the bell, 
the lever receiving an oscillating motion from the escape- 

18, 19, are connected to the piston rod of the cylinder i and 
to the rod of the rack for the signal apparatus. Crank 18 
is provided with a roller 20 and crank 19 with a tappet 21, 

ment wheel so as to cause the striker to strike the bell twice so that on the upstroke of the piston of cylinder i crank 18 


6CQ \\IOA' 

_ 479ae*366 TO uPd.i.K MAIN 

"MF "vewx^fScaeen* ^JH^g- T0 UP NJ ' NE LINC 

- '05 R4<l4a64'T<J UPB.N.P MAIN 

A ^X 


tin 2 Slots Syria/ *;riettty3 in Junction Cabin I Slo t 

- a allSa/M/s " ' S * ' 

- S - - - - // <* " 

/ 12" - 6_ " 

Total JS_Slots 

/-ak f./.fi X gggHJ3aia^ 


; 'ji 






^ Levers for 64 Syna/s & 15 Slots 
31 " 60 Points &SZF:f>. Locks 

3 ' Point Controls 

9 Spare Spaces 
67 Lever Frame 

'S- 4 2 7 (continued). Howrah. 



Fig. 428. Electro-Pneumatic Gates. 

imparts an upward motion to crank 19, whereby the rack- 
rod and weight 10 are raised, while during the downstrokc 

of the piston 20 moves away from 21 so as to leave the 
weight and rack in the raised position. To the crank 18 is 
connected a lever in combination with circuit closing contacts 

22, 23, 24, and with the crank 19 is connected a lever 
operating together with circuit closing contacts 25. The 
downward motion of the barrier is initiated by the contact 
lever 26, situated at a distance, being moved by hand from 
the lower to the upper contact, thereby closing a circuit from 
battery 27 through lead 28, releasing-magnet 17 of the bell 
signal, lead 29, contacts 23, 22, lead 30 back to battery. 27. 
The electro-magnet 17 in attracting its armature 16 sets free 
the signal mechanism, during the ringing of the bell the 
weight 10 gradually descends, thereby causing lever 19 to 
close the contacts 2.15. By this means a branch circuit is 
closed passing from lead 28 to lead 31, contact 25, lead 32, 
valve magnet 3, lead 33, lead 29 and also through contact 

23, 22 and lead 30 back to the battery. The electro-magnet 
in attracting its armature moved the double valve at 5 so as 
to close the communication of the lower end of cylinder i 
with the atmosphere and open the communication between 
the underside of the piston and the supply pipe to the pres- 
sure fluid. The piston is thus moved upward and causes 
the barrier to move downward into the closed position. 

Fig. 424. Earl's Court East Box, Metropolitan District Railway (Sec p. 233). 




The principal " All-Air " system is the Low Pressure 
Pneumatic, in which the points and signals are operated by 
air compressed to a pressure of 15 Ibs. per sq. in., and the 
valves of the point and signal motors by air compressed to 
a pressure of 7 Ibs. per sq. in. 

Such arrangements, of course, necessitate the laying of 
a large number of pipes. There is one main pipe, about 
2ins. diam. inside, throughout the installation. Each point 
lever requires two pipes, iin. diam., one to move the points 
from normal to " over " and another from the " over " to 
normal position, and two more pipes, also in. diam., for 
the "Return-Indication," one for the "over" position of 
the points and one for the normal. Each signal lever re- 
quires a pipe for lowering the signal, a second for restoring 
the signal to danger and a third for the "Return-Indication" 
to show that the signal has gone to danger, no indication 
being given when the signal is pulled off. 

The number of pipes can be reduced in most cases by the 
use of the plug valve illustrated by fig. 454. 

The British rights in this system were acquired by and 
have been developed and improved by the British Pneumatic 
Railway Signal Co., Westminster. 

The first installation in Great Britain was at Grateley, 
on the L. and South Western R., and it included the auto- 
matic signals already described between Grateley and 
Andover. This installation was opened in the summer of 
1901, and was followed by a larger one comprising two 
signal-boxes and a ground frame at Salisbury (opened 
November, 1902) and another at Staines, opened in the 
spring of 1904, and subsequently the four lines between 
Woking and Basingstoke and the Clapham Junction widening 
were equipped. Another important installation is on the 
Great Central R. between Ardwick Junction and Newton. 

Grateley, L. am] South Western Railway. 

There are 72 levers in the frame at Grateley, which are 

3'm. centres or 24ft. over all ; but by the working of facing 

point locks with the facing points, and the actuation of more 

than one signal by the same lever through " selection," 

greater economy still of space is obtained. Of the 72 levers 
in the frame 54 are in work for 46 signals, 17 sets of points, 
and 6 facing point locks, which, with 18 spare levers, would 
require under ordinary conditions 87 levers, which at sin. 
centres would occupy a space of 3&ft. 3ins., or an increase 
of space of practically exactly 50 per cent. 

The diagram, fig. 431, shows the lines and the points and 
signals worked from the cabin. The signal-box is on the 
up platform and there are up and down main lines, an up 

F'g- 43 2 - 
Outside of Signal Box at Grateley, L. and South Western R. 

loop, the Amesbury branch, and sidings on the up and down 
sides of the line. Fig. 432 is a view of the exterior of the 
box, and fig. 433 one of the interior. 

The latter will give an idea of the type of locking frame 
used. The " levers " take the shape of handles, which are 
pulled out when the corresponding point or signal has to be 
worked, and which are painted with the usual distinctive 
colours. The interlocking is of the usual tappet form, and 
is in the front of the frame and, therefore, well in sight 

I, 4, 21, 22, 23, 24, 25. 37, 38, 39,40,47, 48,43, 50,Sl,S2,i3,S4 spare 

Kig. 431. Diagram of Grateley Station, London and South Western Railway. 

91-V , TiMWjfturv 


?7 Levers for 46 Signals. '7 Le*ers& FactwFbinl Loc^S 
>8 spare lews Total ^2 Lexers 









No. 5 is used for holding points No. 32, and when 
lied over is locked in tbat position as long as a train 


pulled over is locked in tbat position as long as a train 
stands on the main between Nos. 4 and 6 signals, and 
b trc circui No wil 

No. 5 will release 

sans on e man eween os. a sgn 

th points No. 32 by track circuit. No. 5 will 
signals Nos. 4, 6, 1, 16, 24, 29, but will be left 
tree of all trailing points. 

-* 15 . 


'.13 IS.28.33.3+.3S.36.37.44.S6.57S8.59.64..SPACES. mtu> ^ mnm u ant 

Fig. 434. West Cabin, Salisbury. 

No. 61 is usad for holding points Nos. 24 and TO, and when pulled over is locked in that position as long as a train stands on the main fnuTea, i cvcp f 
between Nos. 62 and 63 signals and the points No. 24 by track circuit. If points No. 39 are reversed, No. 24 is not held by No 61, so that move- " 
ments through Nos. 24 and 26 can take place. No. 61 will release signals 27, 48, 52 (53, when 50, 51 over), 56, 62, 63, but will be leil free of 
all trailing points. 



SELECTED BY /V9J9 POINTS 202)22. J/.J2.H J*.4J.4S.4fi.5CS 


ig- 435' East Cabin, Salisbury. 

The gasoline engine and the air com- 
pressors are fixed in a shed outside the up 
loop siding-. The air is pumped into the 
reservoir A (fig-. 440), from which a main 
is run throughout the whole length of the 
installation, and from this main pipe there 
are branches to each point and each 
signal. There are also branches from 
each lever in the signal box to valves con- 
nected to the main. 


This installation has several novel 
features which render it more complete 
than any other low pressure installation in 
the world. 

Figs. 434 and 435 are the diagrams of 
the lines into and out of the station, 
at each end of which there is situated 
a signal-box being the East and West boxes. These two 
boxes, and the ground frame which is shown en fig. 435, do 
all the work which formerly was done by four mechanical 
machines. The ground frame mentioned is only used during 
the time that shunting is going on at that point. These 
points, although they are 400 yards away, could easily have 
been operated from the East Box, but the ground frame 
was necessary to meet the requirements of the Board of 
Trade. The interlocking between the East box, fig. 435, and 
the ground frame is carried as follows : 

When shunting is to be done a man is sent to the ground 
frame, all the levers of which he finds locked in the normal 
position. He asks the East box by prescribed bell code for 
an " unlock," which is given by pulling over the lock lever 
No. 60 provided for the purpose, and which, on being pulled 
to the "over" position, frees the king lever Xo. i 

\-\if. 430 l.:iy ou for Holt Locking with G. VV R. at Salisbury. 



in the ground frame and locks up all conflicting 
signals and points in the East box. On obtaining 
this release the man at the ground frame is able 
to operate his points and signals, and when the 
shunting is finished and the levers put normal, and No. 
i lever has been replaced by him to its normal position, 
then, and not till then, can the signalman in the East box 
replace his lock lever, get his indication and free his signals 
and points controlled by it. 

The Great Western R. has a connection with the L. and 
South Western R. at Salisbury (see fig. 434). and the points 
in the connecting lines are mutually controlled, bolt locks 
being provided to each point similar to those illustrated by 
fig. 436. A release lock is fixed which has two slides, one 
coupled to the Great Western box by rodding and the other 
operated from the West box by a slide with a motor similar 
to a point motor. In each of these slides a slot is cut, and 
into these fit a lock which in the position shown in the illus- 
tration is holding the Great Western slide. When the points 
may be worked the South Western slide is pushed and the 
Jock can then move from left to right and free the Great 
Western slide. An electrical indicator is coupled to each 

The West box contains a 64 lever frame of the same 
type as at Grateley. Twenty-two of these levers work 30 
signals, 19 are for points and 5 for the bolt locks between 
the Great Western R. and the West box ; in addition to these 
there are 2 special lock levers controlling signals operated 
from the East box, and 16 spare spaces. The East box also 
contains a 64 lever frame, 27 of which work 32 signals, 17 
are for points, 3 special lock levers and 17 spare spaces. 

The compressed air for working the whole of the switches 
and signals is obtained from one power house situated near 
the East box. There are provided here two compressors, 
each more than sufficient to do the work of the whole 
station, one being kept as a stand-by while the other is 
working. A further precaution is taken by having two 
separate powers to drive the compressors, one compressor 
being driven by means of an electric motor run from the 
mains of the Electric Company at Salisbury, the other by 
steam from a boiler in the boiler-house adjoining ; so should 
either compressor break down or want repairing, its fellow 
is ready to take up the work, and should something go 
wrong with the electric cable, there is the steam power to 
fall back upon, and vice versa. The air is compressed to 
30 Ibs. per sq. inch above the atmosphere, and conducted 
through a galvanised pipe to a receiver of 250 cubic feet 
capacity, from there, having been reduced in pressure to 
20 Ibs. a sq. inch, to two more receivers of the same capacity 
situated one at each box, and from these latter receivers, to 
the signal-boxes, switches and signals, by means of gal- 
vanised pipes. 

Staines, L. and South Western Railway. 
The plant at Staines is very interesting on account of the 
economies effected. Figs. 437 and 438 are diagrams of the 
points and signals controlled from the Junction and the East 
signal-boxes respectively. 

Fig- 437- 

Staines Junction, 
L. & South Western R. 

Ground Frame Interlocking. 

Released by No. Locking. 


4-5 2 

Formerly there were five signal-boxes required here, but 
now it will be seen that there are only two boxes and one 
emergency ground frame. But more than this has been 
achieved. The up and down loops have been extended at 
the east end right up to the road level crossing, and the 
gates at the crossing have been efficiently controlled from 
the signal-box. 

The interlocking in the boxes at Staines is as under : 


Staines Junction Cabin Table of Interlocking. 

Name. Released by Locking 

3 From Windsor Home Sgnls 

4 Windsr to Londn Stng Signl 

5 Windsr to Reading Starting 


6 From Windsor Facing Pnts 

7 Ground Frame Release 

8 Down to Up Windsor Ground 

Signal (thro' ib) 

9 To Windsor Trailing Points... 
10 Windsor Line Crossover 

11 Up to Down Windsor Grnd 

Signal (thro' 10) 10 

12 Up Siding to Up Line Grnd 

Signal (thro' 13) ... 13 

13 Up Siding to Up Line Cross- 


10 32 (6 B & FJ (30 35 
when 6 over) (13 15 20 37 
when 6 normal) 

11 13 17 20 32 (10 B & F) 

(22 26 when 16 normal) (23 
when 16 normal 20 over) 

9 22 26 27 (18 when 15 over) 
(23 when 20 over) 

24 2 5 38 

ii 13 15 20 24 32 37 

6 38 

3 14 25 (18 when 15 over) (22 
26 when 1 6 over) (23 when 
1 6 20 over) (39 when 32 nor- 
mal) (36 when 35 over 32 

.14 Down Windsor to Up Line 
Ground Signal 

4 8 17 20 (3 when 6 normal) 
(36 when 32 35 over) (39 
when 32 over) 

... (15 or 19) 10 25 (18 when 15 over) 
1 6 when 
15 normal) 

15 Down Windsor to Up Line 17 8 37 (3 when 6 normal) (36 

Slip Crossing when 32 35 over) (39 when 

32 over) 

16 From London Facing Points 17 38 39 (3 6 whpn 35 vp r) 

17 To London Trailing Points... 4 "3 '6 20 

18 Up Line to Down Windsor or (1301 15) 12 (6 10 14 38 39 when 

Up Siding Ground Signal... 15 over) 

19 Down to Up Line or Down (20 or 21) 22 (16 B & F) (23 when 20 

Siding Ground Signal ... over) (25 when 16 over) (27 

when 1 6 normal) 




[ 1 / c 18 19. & & Spaces 

r. 438. Staines East Cabin. 

No. Name. Released by 

20 Main Line Crossover 

21 Down Line to Down Siding 


22 Down Sidng to Dwn Wndsr 

or Down Reading Grd Sig 

23 Up Line to Down Windsor or 

Down Reading Grnd Sig... 

24 To Windsor Advance Signal 

25 London to Windsr Start Sig 

26 From London Home Signals 


4 8 13 17 21 26 (3 when 6 
normal) (36 when 32 35 
over) (30 when 32 over) 


London to Reading Start Sig 
Down Siding to Down Read- 
ing Ground Signl (thro' 30) 
Down Siding to Down Read- 
ing Crossover 

Down Reading to Down Sid- 
ing Ground Signl (thro' 30) 
From Reading Facing Pnts 
To Reading Trailing Points 
L : p to Down Reading Grnd 

Signal (thro 1 35) 

Reading Line Crossover 

20 26 

21 6 19 (16 B & F) (10 when 16 
over) (5 30 35 when 16 nrml) 
(1801 20) (10 when 16 Vo over) (16 B & 
F 6 10 when 20 over) (5 30 
35 when 1 6 normal 20 over) 
7 8" (9 10 B & F) 
9 7 10 14 (15 B & F) (19 when 
1 6 over) (36 when 35 over 
32 norml) (;q when 32 nml) 
6 20 21 (16 B & F) (10 when 
16 over) (5 30 35 when 16 
6 30 33 35 (19 when 16 nrml) 




27 33 35 (3 'he.n 6 over) (22 
26 when 1 6 normal) (23 
when 1 6 normal 20 over) 

36 Down (o Up Reading Grnd 

Signal (rhro' 35) 

37 Reading to Londn Start Sig 

38 Reading to Windsr Strt Sig 

39 From Reading Home Signals 

1,2, 28, 40 Spare. 

3 4 8 33 

27 30 32 35 

35 36 (32 B & F) (37 when 32 

over) : (38 when 32 normal) 

27 30 33 39 (3 when 6 over) 

(22 26 when 16 normal) (23 

when 1 6 normal 20 over) 

(31 or 35) 34 (10 when 35 over 32 nrml) 

(13 15 20 when 32 35 over) 

(16 (32 B & F) when 35 over) 

17 15 8 (3 when 6 normal) (34 

when 32 over) 

7 q 16 (34 when 32 normal) 

(18 when 15 over) 

16 35 (32 B & F) (13 15 20 

when 32 over) (10 over 32 
nrml) (18 when 15 over) (25 
when 32 nrml) 

Sfaines East Cabin Tnble of Interlocking. 

2 I 

p Line or Up Line to 
Loop Home Signals 
L'p Loop Starting Signal 
Up Main Starting Signal 
L'p Advance Signal ... 
Up Line to Goods Signal 

8 Up Siding to Up Main or Up 

Loop or Goods Grnd Sig... 

9 Up Sdng to Up Mn Crossovr 

10 Up Siding Catch Points 

11 Up Main to Up Siding Grnd 


12 Down Line lo Down Siding or 

Up Main or L'p Loop or 
Goods Ground Signal 

13 Crossover Poinls 

14 Up Main to Up Loop Cross- 


15 L'p Main to Up Siding or 

Down Line Ground Sig ... 

16 L'p Loop to Gds Slip Crossvr 

17 Up Loop or Goods to Up Sid- 

ing or Down Line Grnd Sig 

21 Goods to Up Line Grd Sig... 

22 Goods or Up Loop lo L'p Line 

Slip Crossover 

23 Goods to L'p Line Slip Cross- 


9 13 23 (14 22 B & F) (16 when 
14 over) (22 when 14 norml) 

22 16 17 23 (10 14 B & F) 

ii 14 15 22 29 (9 13 B & F) 

(22 23 B & F) 
14 16 (21 9 13 (23 B & F) 
when 23 


(9 or 10) ii 14 (16 22 23 B & F) 17 (22 
(21 when when 9 over) (23 when 16 
23 over) normal) 

2 7 10 13 

9 14 

9 48 '4 

(13 or 25) 27 (14 22 23 B & F) (15 when 
(21 when 13 over) (16 B & F 17 when 
13 23 over) 13 14 over) (22 when 13 

over 14 nml) (23 when 13 14 

over 16 nrml) 

2 7 9 26 31 (32 when 28 nrml) 

4 8 10 ii 15 

(ii or 13) 4 14 (12 when 13 over) 

3 (2 when 14 over) 

(loor 14) 3 8 (16 22 23 B & F) (12 when 
(13 when 13 14 over) (21 when 16 
14 over) over) 

23 (16 B & F) (17 when 16 over) 

4 (2 when 14 norml) (8 when 
9 over) (12 when 13 over 14 
normal) 29 

22 2 3 (8 when 16 normal) (12 
when 13 14 over 16 normal) 

No. Name. 

25 Down Siding Slip Crossover 

26 Dwn Lp to Dn Line Crossvr 

27 Down Siding to Down Line 

(jround Signal ... ... 

28 Down Line to Down Loop 

Crossover ... ... ... 

29 Gate Lock 

30 Down Loop Starting Signal... 

31 Down Main Starting Signal... 

32 Down Line or Down Line to 

Down Loop Home Signals 

i, fi, 18, 19, 20, 24 Spare. 

Keleased by 


3 3 2 

'33' (j 2 when 28 normal) 


j 22 32 
26 25 (2.S B & F) 

I 5 2(J 28 

25 2q (26 28 B & F) (13 26 
when 28 normal) 

Woking to Basingsloke, L. and S. W. R. 
The section between Woking and Basingstoke is 23 miles 
in length and has four lines throughout. There are 1 1 
signal-boxes and i ground frame, and particulars of the 
points and signals they \vork were given in Chapter XVI 1 1. 
Automatic or semi-automatic signals are provided through- 
out, the sections averaging 1,500 yards in length. There 
are three power houses : at Basingstoke, at \Yoking and at 
Fleet respectively. 

Ardivick-Neii'ton, (treat Centra] Railway. 

The signalling between Ardwick Junction and Newton 
has been carried out in connection with the Guide Bridge 
widening. It is perhaps the busiest section of the Great 
Central R., being close to the Manchester terminus. In 
the short distance of six miles there are 14 signal-boxes, 
most of them working junctions. Advantage has been 
taken of the new requirements of the Board of Trade as to 
the position of facing points, and it may be safely assumed 
that had mechanical interlocking under ordinary conditions 
been laid down, the number of signal-boxes that would 
have been necessarv would have been nearer twenty. 

Fig. 439 is a diagram of the lines equipped. 

The number of levers and of the points and signals they 
work are shown in the summary given in Chapter XVI 1 1. 
The power house is situated near Guide Bridge East box. 
It contains three vertical boilers and two Class A Duplex 
Ingersoll-Sergeant air compressors, double acting, with 
automatic governor. 

Clupham Junction Widening. 

This important work on the L. and South Western R., 
and which affects five signal boxes, is being signalled by this 
svstem. The frontispiece is an illustration of one of the 
bridges of signals. 

Mechanisms of the System. 

Fig. 440 shows diagramatically the operation of tho 
points and signals. The handles in the frame A i are pulled 
out for a distance of 2ins., and as this is done the operating 
bar A 2, which contains two slots A 3, A 4, actuates the 
interlocking by means of the slot A 3, and opens valve A S, 
which allows a supply of air at the reduced pressure of 7 Ibs. 

K 2 





Fig. 439- 

Diagram of the Low-Pressure Pneumatic Installation on the Great 
Central R. between Ardwick and Newlon. 




64 Levers 

36 Levers 


I 4f 600 Yos. 

1) /, 

28 Levers 



to pass either of the pipes A 10, A 11, according as to 
whether the points are normal or reversed, to the relay 
valves A 12, A 13, and when either of these are opened, air 
at 15 Ibs. pressure is admitted into that side of the piston 
in the cylinder A 14, according to whether valve A 12 or A 
13 is open. As shown in fig. "440 the points are set for the 
right hand road, so that when lever A 2 is pulled, air is 
admitted into cylinder A 14, which pushes the motion plate 
A 15 from right to left, and a pin in the switch-bar travelling 
along the slot A 16, draws the points over, so that they He 
for the left hand road. 

The motion plate travels a short distance, but the slot 
in it produces no effect upon the switch-bar. This motion 
disengages lock A 23 from the locking bar. The further 
travel of the motion plate pulls over the points, and the final 
part of the motion plate locks the points in their new position 
by the lock A 24. The notches in the locking bar are 
differently cut, so that the locks A 23 and A 24 can only 
enter the correct notches. When the motion plate has com- 
pleted its travel, valve A 18 is opened by means of slot A 17 
in motion plate and allows air to pass through pipes A 19, 
A 20 into either of the indicator cylinders A 5, and the effect 
of this is that the pin A 6 or A 7 (according as to whether 
the points are reversed or normal), which has travelled a 
short distance when the lever A i was pulled, is forced by 
the air in the indicator cylinder to complete its travel, and 
in so doing the lever itself automatically completes its move- 
ment, and not only indicates to the signalman that his work 


is done, but the interlocking tappet finishes its travel, and 
so frees its corresponding signal lever. 

The signals are worked in practically the same way as the 
points, except that there is no "Return Indication" for the 
" off " position, the indication only being given for the 
signal at " on." 

All running signals arc protected by " Track-Circuits," 
which are provided throughout. These signals are also re- 
placed automatically, and independently of the signalman. 
This is done by automatically replacing the lever itself and 
making the lever restore the signal. Consequently the lever 
and signal coincide, which is certainly desirable, and is ac- 
complished in the manner illustrated by fig. 441. 

When a train passes over the insulated joints at / of the 
1 Track-Circuit " the track battery t is short-circuited, 
which de-energises the relay d and causes its armature to 
fall, and so breaks the circuit 5 s 2 at e. The magnet m is 
thus de-energised and the cut-off valve g opened and the 
air is allowed to pass from the main pipe h to a replacing 
cylinder h. In the slide a there is a second slot Z and in this 
travels a pin coupled to the piston in the cylinder k. When 
the signal was lowered, the pin was forced into the bottom 
of the slot so that when air is admitted to the cylinder k 
the piston is raised and this causes the piston in k to give a 
left-to-right motion to the slide a, but only for the usual two- 
thirds of the travel. This acts in a similar way to the 
operations of the signalman (closing pipe b, opening b-, 
raising the signal, etc.). The " Return-Indication " comes 

/3 r Tapper #ifh srdinory i 

f--\- ^ / *J 

_ \t/<,, K/M s//rf 

** Supply f'pe A 9 

Fig. 440. Diagram showing operation of Points and Signals by the Low-Pressure Pneumatic System. 


XL. Fig 

ig 441. Replaced Signals. 
a- /-I 

into the cylinder a 2 and completes the stroke and the pin in 
the slot / travels along the horizontal portion of the slot. 

In several cases one lever will operate one of two sig- 
nals, which one depending on the position of the switch 
leading to the track for which these signals give permission 
to run on. The method of doing this will be clearly seen 
from the diagram, fig. 442, relating thereto. 

The indicator selector valve which is placed at the points, 
and worked off the motion plate, is similar to the ordinary 
switch indicator valve, with the exception that it is provided 
with extra ports, and in one position of the motion plate the 
low pressure operating pipe A is connected, by means of 
these extra ports, with the low pressure pipe operating 
diaphragm for signal i, and when the points are moved 
over, the pipe A will be connected to the low pressure pipe 
operating diaphragm for signal 2. 

Let it be assumed that the points in the position shown 
in the diagram will be controlled by signal i, and when the 
points are reversed, by signal 2. 

On reversing the lever in the frame, the low pressure 
air is admitted into pipe A, it travels through the indicator 
selector valve as shown by the dotted line to the diaphragm, 

and operates valve i, so admitting high pressure air to the 
cylinder operating signal i. On the signalman putting his 
lever to the normal position, pipe A is exhausted, and low 
pressure air admitted through pipe B to the diaphragm i a , 
so admitting high pressure air to cylinder i, and putting 
the signal to danger. The signals having gone to danger, 
the same high pressure air is free to go through the indica- 
tion valve i, and then through the cylinder 2 and indication 
valve 2, back to the cabin, which completes the stroke of 
the lever. 

The object of making the indication pipe go through both 
the cylinders in series, is to insure that both signals are at 
danger before getting the indication ; this, of course, is 
obviously necessary, or the indication would be of no value. 

On the Woking-Basingstoke section the traffic does not 
require some of the signal-boxes to be open continuously, 
but if they were closed the advantages of the automatic sig- 
nalling would be destroyed, as some of the sections would 
be longer than others. 

It is an easy matter to arrange a power-worked signal 
so that it will, go to danger automatically, but it is by no 
means so easy to pull such a signal "off" again indepen- 
dently of the signalman. This, however, is successfully 
accomplished in the Low Pressure Pneumatic System by 
setting apart one lever in the machine called the " King- 
Lever. ' ' 

When the box has to be closed all the running signal 
levers are pulled over, and this action releases the " King- 
Lever," which, when pulled over, back-locks the signal 
levers. The electrical and other connections are as shown 
by fig. 443. 

On a train entering the section in advance of the signal, 
the electric relay 3 is de-energised and the circuit a b c 
broken at the contact e. This de-energises the electro- 
magnet p , opens the cut-off valve g and admits air to the re- 
placer cylinder k on the signal lever B. But as the lever is held 
by the back-locking on the " King-Lever " the signal is put 
to danger by the air passing through the valve g, operating 
the cut-off valve n, and the communication between pipes o 
and m cut off, and pipe m under the diaphragm q exhausted, 
and thereby the signal allowed to go to danger. At the 
same time the air, passing through the electro-pneumatic 
cut-off valve g, acts on the relay s and admits air from the 
pipe o which, operating on the slide valve t, closes the 

Fig. 442. Diagram of Selected Signal. 





Fig. 443. Automatic Station Signal Working. 

opening v and passes along pipe 70, operates relay x and 
admits air to the upper end of the cylinder r, thus ensuring 
the arm being put to danger. When the train has passed 
out of the section, the electric relay 3 is energised and 
circuit a b c is made through the electro-magnet p, which, 
by closing the valve g, exhausts the air from under the 
diaphragms of the cut-off valve n and the relay s admits air 
from the pipe o to the pipe m and so to the relay q at the sig- 
nal and at the same time cuts off the air supply through 
relay s, the pipe w operating the relay x is exhausted, and 
the air on the upper end of the cylinder r is also exhausted 
and the signal is again cleared automatically. 

When the box has to be re-opened the " King- Lever " 
is first replaced to normal, which releases the running 
signal levers. 

Distant signals in the Low Pressure Pneumatic System 
are not provided with levers, but are automatically lowered. 
A contact maker is fixed on the arms of those stop signals 
to which the distant signal applies, and when they all are 
" off " a circuit is completed to the distant which lowers 
it. Any of the signals being restored to danger causes the 
distant to go " on " again. 


ig. 444. Holding Roads. 

In fig. 444 is illustrated an arrangement whereby the road 
is held even when the signal has been put to danger. This is 
particularly applicable to junctions and other places where 
the protecting signal is some distance from the points or 

The lever-slide u operates the points b, and in the slide 
are two notches c c- corresponding to the normal and reverse 
positions of the lever. When the signals d d are " off " the 
points are locked by interlocking, but as soon as the signals 
are at normal the points are free. The lever is, however, held 
in this way: "Track-Circuit" sections exist between the 
joints e e e e e e, the battery of which holds the relay / closed. 
When a train enters the relay is de-energised and the contact 
f falls away and the circuit to the cut-off valve g is broken 
and it is opened so that air enters and raises the lock h into 
the notch c or c-, according to the position of the lever. The 
latter cannot therefore be moved until the train has passed 
over the section, when the track battery again energises relay 
/, and the air supply being cut off the lock h falls. 

I ' 2 34 

9 10 II 13 14 15 16 17 

28 29 30 31 32 

'fl/oc< Irttrrs 2'aport -REVERSE 

''g 445- Push-Button Signalling at Wath Concentration Sidings : Great Central Kailwav. 



Wath Concentration Sidings. 

These sidings of the Great Central Co. at Wath in York- 
shire arc operated on the gravitation principle, whereby there 
is one " hump " for the loaded wagons and another for the 
empties. The wagons arc pushed up from the reception 
sidings to the " hump," there detached and run by gravity 
into their respective sidings. 

The points giving access to these sidings have to be 
moved over very quickly, as it is often necessary to pull over, 
or reverse, a pair of points between two wagons. At the 
celebrated gravitation sidings at Edge Hill and Aintree the 
points are worked by rodding from a signal-box in the usual 
way. The Great Central Co. have, however, gone in for a 
new departure, so far as this country is concerned, inasmuch 
as that the points are operated by power from a " push- 
button " machine. 

The signalling has been carried out by the British Pneu- 
matic Railway Signal Co. , and the points are operated by low 
pressure compressed air, the valves controlling which are 
opened and closed electrically by currents set up through 
switches in the "push-button " machine. 

Four signal-boxes have been provided, two of which 
Elsecar Junction and Wath Station control the entrances to 
the yard from the main line. Two others have been fixed in 
the centre of the sidings. One of these, B, on the north side, 
controls the lines for " fulls," and the other, C, on the south 
side, those for empties. The former box has one of the usual 
pneumatic frames, with 7 working levers and 5 spare. The 
7 levers work, by the usual low-pressure system, 5 signals, 
four pair of switches with locking bars, and i turntable bolt 
lock. Box C contains a similar frame of 12 levers with 6 
spare. The 6 working levers operate 5 signals and 4 pair of 
switches with locking bars. In addition they have each a 
" push-button " machine with 33 buttons, each box having 
4 spare buttons. 

Fig. 445 illustrates the "push-button" machine,- "but 
before its details are given it is necessary to explain that each 
group of sidings fans out from one line into two ladders. The 
splitting point for the two fans is worked from the signal-box 
by what is known as the " King " button K in fig. 445. 
There are two buttons for each point, and when all is normal 
the upper row of buttons are in and the lower row out Nor- 
mally the points worked by the " King " button lie for the 
northerly ladder, on which s;ide are all the points worked by 
even-numbered buttons, all of which are coloured black. The 
" King " buttons are coloured differently to the rest, as the 
upper is red (shown hatched in the illustration) and the lower 
black. This indicates that to get a red button the " King " 
red button must be in, but to get a black button the black 
" King " button must be in. " King " buttons control the 
electrical switches of the other buttons so that the latter must 
correspond as outlined above. 

Fig 446. Indicating Signal for Wath Concentralion Sidings. 




Self-indicating Cylinder. 

\Yhen the lower button a is pressed in, the switch b 
centred at r, breaks contact at d and makes contact at e and 
the upper button / is pushed out. An electrical circuit is thus 
completed for reversing the points controlled by the button 
from " normal " to " over." 

In the spare between each ladder 8 cast-iron boxes are 
placed, each of which contains 4 electrically-operated valves 
one for each point, and therefore one box for every 4 points. 
These valves admit air to the one side or other of cylinders, 
which passes thence to the motor at the points, which motor 
is x>f the usual low-pressure type. 

Above the push-buttons are two rows of indicators g g" 
which normally are held hidden. " Track-circuits " are laid 
down throughout the h) ciders and each point protected. As 
soon as the switches at the points are fully " home " in one 
position or the other the indicator is hidden, but directly the 
' Track-circuit " is broken by the points being opened a red 
disc appears behind the screen h. The signalman can there- 
fore, by watching the indicator, see that the points have 
responded to the button. 

Each point is protected by " Track-circuits " laid in to the 
fouling points, and as long as a wagon is not clear the indi- 
cator of the points that are fouled appears. The signalman 
may thus follow the movements of the wagons, and should 
any have stopped foul he must make his arrangements accord- 
ingly. This provision is exceedingly useful in foggy 
weather and also at night. 


Fig. 4-)t>a. Locking Arrangement for Indicating Signa!. 


-is i 



Fig. 447. 

Diagram of Piping for Indicator 
Signal, Elsecar Junction, G.C.R. 

Pneumatically Operated Route Indication Signal. 

On page 67 of Mechanical Railway Signalling there is 
described Annett's Route Indicating Signal. This is 
mechanically worked. The British Pneumatic Railway Sig- 
nal Co. have now erected such a signal (operated by their 
low-pressure pneumatic system) at Klserar on the Great Cen- 
tral R. at the western entrance to the Wath concentration 
sidings, and another at Moor Road at the eastern end of the 

Fig- 44 6 illustrates the signal, on the front of which, 
below the arm, is fixed a frame carrying slides bearing 
numbers. This particular signal carries ten slides, but only 




those 1-9 inclusive are connected. The slides are hidden from 
view normally and held by latches, there being a latch for 
each slide. 

This signal will lead to any one of nine roads, and when 
the road is " made " for a movement the corresponding latch 
is withdrawn and frees its slide. The slide will not fall, how- 
ever, until the signal is- lowered, as all the slides are held up 
by the frame b, the top b of which is under all the slides. 
There is also a shutter c which covers the opening into which 
the slide falls, and this shutter has to be raised. This is 
done simultaneously with the lowering of the slide and the 
movement of the signal arm, as when power enters the 
cylinder d at the foot of the signal the operating rod e is 
raised. This raises the lever c 3 , centred at c' J , and lifts the 
shutter c and at the same time the lower end of lever b- is 
raised and the upper end lowered, which brings down the 

coupled by the rod /i 3 to the piston in the cylinder /. As the 
red /i 3 is raised (and lowered in the case of the suspended 
cylinders) the latch g is withdrawn from under its slide and 
the latter is free to fall when the frame (b fig. 446) is lowered 
as the signal arm is pulled "off." 

Fig. 446/7 gives details of the cylinder / which had better 
be studied in connection with fig. 447, which is a diagram of 
the pipe connections. 

From fig. 447 it will be noticed that one road fans out 
into nine, for which, ordinarily, there would be required a 
signal with nine arms, and in the locking frame nine levers 
for the same. But as here described there is only one arm 
and one lever No. 18 which has to be " led " by No. 14 or 
15 or 16 or 17, No. 14, for example, being selected by No. 35 
points and No. 16 by No. 39 points. 

rod b 3 coupled to the frame b. When the signal is restored 
the screen falls and the frame rises, carrying back the slide, 
which movements are assisted by the weights c 4 and b*. A 
lamp is provided to illuminate the slide at night. 

The latches are withdrawn pneumatically by power con- 
trolled by levers (slides) in the usual locking frame. Fig. 
44h gives details of the cylinders, motion plates and latches. 
There are nine cylinders, /, five on one side and four on the 
other of the post and two of each set are above and three 
below the latches. There are nine latches g, the point g- of 
which rest under the slide. Secured to the latch is a pin, on 
which is a roller working in a slot /;- in the motion-plate h 


Let it be assumed that a train has to be admitted into Xo. 
i road. For this Xo. 14 lover with Xo. 35 points normal 
has to be pulled. This done air is admitted to reverse pipe i, 
through the selector 2 into pipe 3 and then to port h (m fig. 
446/7) of cylinder / (figs. 44(1 and 4460), raising the piston r 
and the rod /i 3 (iig. 446(1). \Vhen the rod has been raised to 
release the slide the piston clears the port ;' (n fig. 466/7) and 
air then flows along pipe 4 to the plug valve 5 and thence by 
pipe 6 to the reverse diaphragm relay in cylinder d (fig. 446). 
When, now, Xo. 18 lever is pulled air is sent through pipe 7 
into the lower side of cylinder J, so raising the operating rod 
e, lowering the arm and displaying " Xo. i "on the screen. 

When the signal has to be restored, Xo. 18 lever is put 
back, air is admitted into pipe 8 (fig. 447), which opens the 
normal diaphragm relay in cylinder d and forces down the 
piston, so lowering the operating rod e and raising the arm 
and the frame b and, at the same time, the shutter c falls. 
\Vhen the movement is completed the air flows through pipe 
9 (fig. 447) and gives the " Return-Indication," so completing 
the stroke of Xo. 18 lever. This frees No. 14 lever and it 
may be replaced to normal, thereby causing air to flow 
through pipe 10 to plug valve 11 and through pipe 12 to 
cylinder / 3 , thence by pipe 13 to cylinder / 8 , and so on through 
all the cylinders, the last being /-, whence air passes through 
pipe 14 to the locking frame, when it enters the " Return 
Indication " valves on levers Xos. 14, 15, 16, 17. The pistons 
in the last three are already normal, so only Xo. 14 requires 
raising. Similarly the pistons in all the cylinders were normal 
except that in cylinder /, but by going through them all it is 
guaranteed that they are all normal. The piston in cylinder f 
is reversed by air entering through port o (fig. 446/7), and 
thence it flows through port / to cylinder / 4 (fig. 447). 
Plug or Sliding Valve. 

Reference has been made to a plug valve. This is illus- 
trated by fig. 448. By its use nine pipes are led into one com- 

mon pipe at 5 (fig. 447) and one common pipe into four at n. 
Its use may be appreciated if it is assumed that air from 
cylinder f 1 enters, through pipe 14, at port s (fig. 448). This 
would force over the valve t to the right and closing the port 
u and passing through port v to pipe 6. Pipe 15 from cylinder 
f is joined at ;/, and when air passes therein the valve / is 
forced over and closes the plug s. 

This reference to the plug valve allows for mention of a 

Fig. 448. Use of Sliding Valve. 

further economy which is possible very often. Where two 
signals, as at A B in fig. 449, are conflicting and only one can 
be " off " at the same time, only one pipe instead of two is 
required to put these two signals to danger, and only one 
pipe instead of two for the " Return-Indication." The pipe 
16 comes from lever A up to the plug valve 17 and the pipe 18 

Fig. 440. 

from lever B and thence by one common pipe 19 to signal B 
and thence to signal A, putting that one that is " off " to 
danger and guaranteeing that the other is " on," and the 
" Return-Indication " for both levers is effected by one pipe 
20, as only one indicating valve can be down. 

l-Disengoger for Road 
5-Disengager for fioad V 
ll-D/sengager for Lever 2 
fur Lever 4- 

15- Trajecteur for Road M 
22 - Trajec teur for Garages Cen frvux 

10. 40. 41. 42.45. 44. Spaces 

Fig. 450. Diagram of Lines. Low-Pressure Pneumatic Signalling at Ermont ; Northern Railway of France. 



Ermont, Northern of France Railway. 

So far as the locking frame and the point mechanism arc 
concerned, and, to a certain extent, also the operation of the 
signals, this installation is identical to those at Gratelcy, 
Salisbury, and Staincs, already described. To British and 
American readers the novelty lies, firstly, in the modification 
of the system to suit French ideas and practice, and, 
secondly, to the introduction and operation of a system of 
trajcctcurs or route levers, and, thirdly, the adaptation of 
pneumatic work in connection with the Aubinc treadles for 
zone locking. 

By the adoption of the system of trajecteurs, which is the 
invention of M. A. Cossmann, Chef des Services techniques, 
and his assistant, M. M. E. Despons, the switches or series 
of switches for a movement are made by one lever. Put 
briefly, it may be said that the lever slide in the locking 
frame, when pulled half-way, moves over in series the 
switches for the required movement, and when all are over 
and locked the "Return Indication" is received in the cabin, 
which automatically completes the stroke of the lever, and 
this frees the lever through the interlocking for working the 
signal for the switches set up. Should anything get " hung 
up " at the points, the " Return-Indication " not being re- 
ceived, the signal lever cannot be -pulled and the signalman 
would know that something was wrong. 

Another feature in the Ermont signalling as compared 
with English practice is the use of one common signal for 
all directions from which a movement can be made, i.e., if 
from any one position a movement can be made to any one 
of five directions there is only one signal instead of five. 

A further feature is the use of invcrseurs for crossing 
movements, of which there are two one, No. 18, for move- 
ments from left to right, and the other, No. 19, for move- 
ments from right to left. One of these is pulled after the 
trajecteur lever has completed its stroke, and it lowers the 

The features then are : (a) trajecteur levers for the 
operation of a series of points concerned in a movement ; 
(b) two inverseur levers for working the signals for crossing 
movements ; and (c) one common signal for leading from 
one position in any direction. 

Fig. 445 is a diagram of the lines and installation. There 
are four main lines two (marked III. and IV.) between 
Paris and Pontoise and two (marked V. and VI.) between 
Paris and Valmondois. There are sidings (Garages de 
Gauche) outside the line from Paris to Pontoise and be- 
tween the main lines (Garages Centraux). 

The frame in No. 2 box consists of 38 working levers and 
6 spaces. As each point and each signal can be operated by 
one or by several levers, it is not practicable to distinguish 
the points and signals by numerals as is usual, so letters are 
employed in the diagram. 

All stop signals are square-shaped and distant signals 
diamond-shaped, whilst warning signals arc circular. The 
last stand normally " off " and can be thrown to " danger " 
when required. Numbers are fixed against all points and 

signals, and these indicate the levers that will operate them 
when pulled in proper sequence. 

For the locking up of a series of switches in a zone 
governed by a home signal a system of Aubinc treadles is 
used. One treadle is installed at the entrance and another 
at the exit of the zone to be protected during the passage 
of a train. Referring to line III. from Paris to Pontoise, 
when No. 2 lever is operated it works in series, ist, the 
treadle A 2 ; 2nd, the treadle A 1 ; 3rd, the signal A by wire 
connection from the treadle A 1 ; 4th, the signal A bis . When 
the train passes over the entrance treadle A 1 it replaces the 
signal A to " Danger " mechanically and at the same time 
sends a current of air to the signal lever 2 and pneumatically 
locks it in the reverse position, thus holding the interlocking 
on the switches until such time as the train passes clear of 
the exit treadle A 2 . The lever is then free to be put back 
part way by the signalman, and the " Return In- 
dication " received automatically, restores the lever, 
ensuring that the two treadles and the signal have 
been put to their normal position. Similar treadles are pro- 
vided also in lines IV., V. and VI., worked respectively by 
levers 4, 6 and 8. 

Should it be required not to send the train forward to the 
next section over the exit treadle A 2 , but to dispose of the 
train on to another running line or siding, a separate lever 
is provided, say No. n, which, when operated, takes the 
"Lock of Transit" off No. 2 signal lever, allowing that 
lever to be put back to normal, thus freeing the switches, 
and after the shunt movement has taken place No. n lever 
is put back and restores the treadle to normal, and, the 
return indication being received automatically, places the 
lever to normal. Lever No. 12 acts similarly for line IV. 

No n releases No. 25 lever, which turns signal K and 
allows a wrong road movement along line III. to the 
station. Before No. 12 lever can be moved No. 3 has to 
be pulled, which puts to danger signal O that normally 
stands " off," and putting signals H 1 I 1 to danger if "off" 
or keeping them at danger. When Nos. 3 and 12 are over 
No. 26 can be pulled, and this turns signal B for a wrong 
road movement along line IV., and No. 29 is released which 
operates signal F for a similar purpose. 

No. 5 operates the disengaging apparatus, at Cabin No. 
i, for the line to Valmondois. 

Levers XYZ work switches in the Garages Centraux. 
Levers 38, 39 bolt the wagon turntables. 

Fig. 451 is a diagram of the connections to a signal 
equipped with an Aubine treadle. The treadle apparatus a 
consists of two parts, the lower being connected to the 
treadle b and the upper to the pneumatic cylinder. The 
signal is connected by the wire d to the lower part of the 
mechanism. When the signal has to be lowered, air is 
admitted to the left side of the cylinder so that the rod 
h is turned from right to left and both parts of the mechan- 
ism turn. There is a recess in the lower plate of the 
mechanism into which the weighted arm b 2 drops, and this 
raises the bar b. The wire d is also pulled so that the signal 



Fig. 451. Aubine Treadle. 

Fig. 4510. Aubine Treadle and Bar. 

is lowered. When a train passes over the treadle it is de- plate of the mechanism is now returned to normal, which 
pressed and the arm b" raised out of the lower plate, so that takes with it the rod f. 

the weight on the signal pulls back the lower plate but the 

Fig. 452. Connections for Disc Signal. 

upper remains undisturbed. Another effect of this move- 
ment is that the rod c- actuates the slide valve e, whereby 
a supply of air is sent to the signal-box and causes a pneu- 
matic lock to enter the lever working the signal, so that 

Fig. 453. Switch Lay-Out. 

the lever cannot be put to normal. The indication box i 
is for giving the "Return Indication" when the slide has been 
freed by the train going out of the section, and air being 
admitted to the cylinder at the other end so that the upper 

Locking bars g are provided to the exit treadles so that 


Fig. 4530. Switch Lay-Out. 

they cannot be operated whilst trains are passing over the 
treadle. Fig. 451^ is a di